Characterizing Nonreturning College Students and Their Departure Intentions

Abstract

Success for 4-year universities is often measured by graduation and retention rates; however, gaps exist in understanding nonreturning students at private institutions. Recent research is helping to build the lexicon of drop-outs, stop-outs, opt-outs, and transfer-outs to inform strategic retention initiatives. Using an action research method, we characterized 1,091 students into designated subclasses and utilized exit interviews, advising notes, and university surveys to theme their departure intentions. Findings reveal transfer-outs to be the largest subclass, with departure reasons being summarized within themes of university experience and fit, health, academic, and financial. Recommendations are made for university administrators and retention offices related to exit survey questions, broadening success metrics, and to focus on specific student characteristic groups.

Keywords

retention student success action research qualitative

For 4-year universities, success is often measured by graduation and retention rates (Millea et al., 2018). Higher education institutions (HEIs) have long sought to identify characteristics to predict student retention and attrition to achieve success (Brasher et al., 1980; Gathers, 1982; Tichenor, 1987). Nonreturning students depart HEIs for many reasons (Hermanowicz, 2006). Yet, the evolving complexity of these reasons has generated a need for both more than one descriptive category and an instrument with a wider scope than a traditional exit survey (Hoyt & Winn, 2004). Building upon the work of Hoyt and Winn (2004), we studied student departure data from a selective, private, 4-year university to expand a singular descriptor of nonreturning students into four categories: drop-outs, stop-outs, transfer-outs, and opt-outs. We aimed to determine whether the success of retention initiatives is constrained by legacy descriptors that classify nonreturning students as a single population. Furthermore, our action research orientation sought to assist practitioners in improving and refining the lexicon of student retention (Sagor, 2000). In doing so, we hope to accommodate the complex intentionality of departure while informing prediction models that could improve measures of retention success.

Expanding Student Readiness with a Retention Focus

In his foundational college retention theory, Tinto (1975, 1993) utilized a sociological lens when developing his model to understand student departure. This seminal theory highlighted individual entry characteristics (e.g., pre-college schooling, individual attributes) as the predictive factors contributing to student departure (Demetriou & Powell, 2014; Tinto, 1975). More recently, the field of student retention has undergone a shift from focusing on individuals’ sociodemographic characteristics as the sole predictors of student persistence, to utilizing student development theories as the method to better understand the relationship between student development during college, including how external forces shape their experience, and student retention and success. This shift can be seen through the suggestion by Demetriou and Powell (2014) that developmentally focused retention policy “may have more to offer than the dominant sociological paradigm” (p. 419). Their work included transitioning from a college-ready model (with the onus on the student to acclimate to and be prepared for the institution) to a student-ready model (with the onus on the institution to be prepared for all students). This type of model and accompanying research offers a perspective in which “positive development occurs through reciprocal and mutually influential relations between an individual and the multiple organizational levels of an environment” (Demetriou & Powell, 2014, p. 427). Student development theories have led researchers to promising outcomes in academic well-being and student retention (Shek & Chai, 2020; Taylor et al., 2017; Turner & McDaniel, 2016).

Furthermore, research on college-student retention has focused on public institutions, leaving much to be understood of private HEIs. In his research on retention at an elite, private university Hermanowicz (2006) suggested student reasons for leaving college as being pluralistic and ill-formed but sharing “generic properties” that can inform retention policy (p. 21). He stated that student reasoning offers less of a “valid ground to leave” and more of an opportunity for cognitive intervention (p. 35). A 2006 study of first-year Latino students at a private university determined that the sociological factors often treated as predictors of retention were not statistically significant (Chowdhury, 2006). From research at a midsized, private university, McPherson (2016) identified “multitheoretical retention levers” to reduce student withdrawals from college (p. iv). More research is needed on the efforts of private HEIs to retain students.

Nonreturning Students

Working Toward Retention

As with most fields, research on college-student retention is evolving. Prior to the 1990s, retention research largely treated nonreturning students as a single category (Hoyt & Winn, 2004). In the 21st century, however, specific issues of financial solvency, university ranking, and brand reputation have become inextricably linked and measured through student-retention initiatives (Delen, 2011). Building the “sustainable university” is of growing interest to higher education researchers and practitioners alike (Adomssent et al., 2007; Amaral et al., 2015; Gal, 2012; Neamţu et al., 2020; Zahid et al., 2017). Higher education practitioners will recognize the sustaining power of successful retention initiatives. In an article presented at the 2005 National Conference on Student Recruitment, Marketing, and Retention, leading retention researcher Tinto (2006) addressed “what's next” in research and practice: “Unfortunately, most institutions have not yet been able to translate what we know about student retention into forms of action that have led to substantial gains in student persistence and graduation” (p. 5). While some have focused on retention strategies that increase engagement (Davis et al., 2013; Zhao & Kuh, 2004), that engagement often treats nonreturning students as a single category. This treatment leads to “inaccurate findings in the prediction of student retention”―predictions that inform misguided campus engagement strategies (Hoyt & Winn, 2004). Therefore, a need exists to operationalize multiple categorizations to aid actions by institutions to improve retention.

Broadly speaking, recent research has enhanced our understanding of nonreturning students. Namely, financial stress, self-reported student loan debt, and sense of belonging represent useful predictors of student retention (Britt et al., 2017; Davis et al., 2019). International and first-generation students tend to experience lower levels of retention than domestic and non-first-generation students (Haverila et al., 2020; Soria & Stebleton, 2012). Gender, ethnicity, and socioeconomic status also interact to affect student retention rates (Keels, 2013). And lastly, while increased access to higher education has contributed to social and economic benefits for historically underrepresented groups, it has also shed light on the issue of student retention and success for these same populations (Thomas, 2016). Thus, it remains important for higher education leaders to differentiate their retention efforts to ensure a campus ready for student development.

Categorizing Nonreturners

In the 1980s, a notable few researchers began exploring ways to differentiate students’ reasons for withdrawing from college through subclasses like drop-outs, stop-outs, horizontal transfers, vertical transfers, and nonpersisters (Brasher et al., 1980; Gathers, 1982; Tichenor, 1987). These early researchers recognized that “one of the most important tasks of retention research is consequently to enable the administrator [to] distinguish between these importantly different types of withdrawal” (Ewell, 1984, p. 11). Recent research is now helping to build the lexicon of drop-outs, stop-outs, opt-outs, and transfer-outs, as well as informing strategic retention initiatives. Stephenson et al. (2020) found that participation in a voluntary survey in their first semester “had notable retention prediction power” for identifying drop-outs (p. 17). Webb and Cotton (2018) proposed increasing one-on-one interactions between students and faculty, as well as peer interactions, as a strategy to reduce drop-outs and increase retention. Inside Higher Ed recently reported on organizational initiatives focused on reenrolling “stopped-out students” across 13 colleges in 12 states following the global pandemic (St. Amour, 2020, para. 2). For opt-outs, James (2020) suggested that nontraditional-student retention can be increased by better preparing them for the difficulty and time commitment required of online learning. Additionally, it was demonstrated by a recent study that students who transfer out typically exhibit lower levels of social integration at their initial institutions (Ishitani & Flood, 2018). Bringing together recent research around a common lexicon can have important implications for campus-ready retention initiatives targeted at a specific departure population and reasons.

Conceptual Framework

In the early 2000s, Hoyt and Winn (2004) reiterated the need to expand the treatment of nonreturning students beyond a single category. We utilized their conceptualization for the framework of our study. “The problem with treating nonreturning students as a single population is the implications of inaccurate findings in the prediction of student retention” (Hoyt & Winn, 2004, p. 396). Their research at a large, suburban, public university in the Rocky Mountain region of the United States expanded the treatment of nonreturning students to four subclasses: drop-outs, stop-outs, opt-outs, and transfer-outs. Three of these categories find commonality in their degree-seeking objective (drop-outs, stop-outs, and transfer-outs), while one (opt-outs) differs in its non-degree-seeking objective. Table 1 includes definitions and characteristics for each subclass of nonreturning student applied in this study.

Table 1.
Subclasses of Nonreturning Students.

Opt-out Drop-out

Definition: Students who leave the institution having accomplished non-degree-seeking objectives
Characteristics: Nondegree-seeking, avocational (i.e., college is hobby to main vocation); attend to earn certificates or some form of knowledge Definition: Students who leave the institution permanently and do not reenroll toward degree-seeking objectives
Characteristics: No plans to return; do not transfer to another HEI; usually motivated by financial reasons, family problems, or academic difficulties; tend to be older than typical students

Stop-out Transfer-out

Definition: Students who leave the institution temporarily with plans to reenroll toward degree-seeking objectives
Characteristics: Do not complete degree within the normal time schedule; health issues; financial reasons Definition: Students who leave the institution permanently by transferring to another HEI toward degree-seeking objectives
Characteristics: Status-driven; accessibility to different programs or program availability driven; financial reasons

Note. Adapted definitions from Hoyt and Winn (2004).

This conceptual framework promotes a richer understanding of the complex intentionality of student departure from HEIs. Specific to our research is a focus on an understudied area of higher education research―elite, private HEIs.

Purpose and Relevance

Given the lack of research available on retention at private institutions (Hermanowicz, 2006) and the need to further understand nonreturning students (Hoyt & Winn, 2004; St. Amour, 2020; Stephenson et al., 2020), we conducted this study with three purposes: (a) classify nonreturning students using the Hoyt and Winn (2004) framework; (b) summarize the departing intentions of nonreturning students, and (c) compare the number of nonreturning students falling into each subclass based on student characteristics (i.e., student demographics).

Method

Design

Based on the suggestion by Altrichter and Posch (1989) that “what's good for the practice is good for the research” (p. 29), we utilized a method that worked for us as practitioners. Our pragmatic paradigm informed our descriptive, historical (Creswell & Creswell, 2018), action research method to understand what is occurring (i.e., undergraduate-student departures). Action research is a disciplined process of inquiry conducted by and for those taking the action (Sagor, 2000; Melrose, 2001). The primary reason for engaging in action research is to assist the actor (i.e., research team members with responsibility for retention) in improving and/or refining their actions (Sagor, 2000). Action research is appropriate if three conditions are met: (a) the focus of the action research project is on the actor's professional action, (b) the actors are empowered to adjust future action based on results, and (c) improvement is possible (Sagor, 2000). These three conditions were met by the present inquiry as (a) the research team members have responsibility for retention, (b) members of the research team have control and influence to adjust future processes/policies in their respective units to this end, and (c) the research site's undergraduate retention to graduation can be improved and has a stated goal to improve. The Institutional Review Board at the research site reviewed and approved the study.

Population and Sample

Patton (1990) posited the use of purposive sampling when the population is identifiable to the researchers. The population identified for the study was degree-seeking undergraduate students who entered the research site from fall 2017 through fall 2020 and who departed the university temporarily or permanently during fall 2017 through spring 2021. Nonreturning students were operationally defined as those who did not complete one or more fall or spring terms at the research site. Utilizing a roster method (Marsden, 1990) the research team exported a list of 1,091 noneturning students from the student information system in June 2021. A total of 69 students who withdrew for medical issues or were dismissed for academic or conduct reasons were excluded from the sample. Students who were deceased at the time the data were exported were also excluded. Table 2 details the population of the student sample for the study.

Table 2.
Demographics of Nonreturning Students Between 2017 and 2020.

Demographic Count (n) Percent (%)

Sex

Female 502 46%

Male 589 54%

Classification

1st-year 300 28%

2nd-year 575 53%

3rd-year 180 17%

4th-year 36 3%

Admission type

1st-year 803 74%

Transfer 216 20%

1st-year transfer 72 7%

Cohort term

Fall 2017 340 31%

Spring 2018 43 4%

Fall 2018 302 28%

Spring 2019 36 3%

Fall 2019 270 25%

Spring 2020 36 3%

Fall 2020 64 6%

Ethnicity

American Indian/Alaska Native 1 0%

Asian 80 7%

Black or African American 62 6%

Hispanic of any race 138 13%

Native Hawaiian/Other Pac-Islander 0 0%

Nonresident alien 97 9%

Race and ethnicity unknown 3 0%

Two or more races 57 5%

White 653 60%

Special populations

Pell-eligible in 1st year 188 17%

First-generation 63 6%

Athlete 3 0%

Fraternity and Sorority members 163 15%

University honors program 42 4%

Data Collection

Goodson and Walker (1991) recommended that “the task of research is to make sense of what we know” (p. 107). Therefore, existing educational-student records and data were used to answer the research purpose(s). Specifically, the research site's student information system (SIS), academic-advising notes, exit survey notes, university-administered surveys, and National Student Clearinghouse (NSC) data were combined to form the richest description (Lincoln & Guba, 1985) and institutional understanding for a student's departure.

Student Information System

During the planning phase of the study, the research team brainstormed and identified student data maintained in the SIS to aid in understanding a student's decision to depart. The research team exported student characteristics from the SIS, which included three types of data best conceptualized as (a) academic, (b) demographic, and (c) campus involvement. Academic data exported from the SIS included variables such as admission type, cohort term, graduation status, admission rating, and term-specific data, including enrollment status, academic-standing status, and degree-seeking status. Demographic data included variables such as sex, ethnicity, first-generation status, and Pell eligibility in their 1st year. Finally, exported variables of campus involvement included athlete status, Greek life affiliation, and membership in a scholar group specific to the research site.

National Student Clearinghouse

The research team then requested all enrollment data prior to May 15, 2021 (including the spring 2021 term and earlier; however, spring 2021 data were not yet available) from the NSC for the 1,091 identified nonreturning students in the sample. The research team merged the NSC data with the SIS data and identified 416 students who attended other institutions after their entering-cohort term at the research site. Some students showed as enrolled at both the research site and another HEI within the United States in the same term; in those instances, the research team considered the student enrolled at the research site. Data reporting standards and the operational definition of enrolled students can be found on the NSC website (National Student Clearning House, 2021).

University-Administered Surveys

Two university-administered surveys collect data from or about nonreturning students to aid retention initiatives at the research site. The first is the research site's student exit survey created and disseminated utilizing an online survey platform, Qualtrics. This survey was sent from staff members within the academic success office to identified respondents starting in April 2020. Respondents were identified as eligible for the survey when they notified campus staff or faculty that they intended to permanently leave the institution. The survey was sent immediately to the student from the staff or faculty member, once identified. The survey was primarily completed by undergraduate students themselves, but at times, staff knowledgeable of a student's experience completed it on their behalf. This action was generally taken when a student was nonresponsive to requests for survey completion. The survey consists of 69 items, and display logic and branch logic were utilized to pose additional questions based on a respondent's answers. Of the 69 items posed to respondents, the answers to four open-ended items and one nominal item were included in analysis. There were 130 exit surveys accessible to the research team, of which 85 responses were from students contained within the sample.

The second university-administered survey is a Qualtrics-based survey of staff and faculty on student retention. It is distributed to selected staff and faculty who have engaged regularly with students. The survey contains 11 survey items with the purpose to provide a space for faculty and staff to notify the academic success office, at any point in the year, when they are concerned that a student may be at risk of leaving—both permanently and for short-term leave. In total, this university administered survey had 209 submissions, of which 106 were included in the analysis because they included data about students within the sample.

Academic-Advising Notes

Advising notes document student interactions, primarily between a student and an advisor, via the SIS. This function allows advisors and other staff to record their notes and provide documentation to others authorized to view these student data. The categories documented are academic discussion, graduation discussion, personal discussion, international student services discussion, and student considers leaving university. For the purposes of this study, we only included data identified by the category of student considers leaving university. Within this category, the following subcategories are utilized by advisors and campus staff: financial issues, leave of absence, and transferring to another school. The documentation of a student discussing leaving the university and the citation of one of the subcategories were used when placing students in the appropriate subclass of nonreturning student and/or student intent. Advising notes were implemented in April 2020, with a total of 46 relating to the sample being exported.

Exit Interviews

Exit interviews were conducted from May 2019 to August 2020. The interviews consist of six questions asked by a staff member in the academic success office to the nonreturning student either verbally (in-person or by phone) or via email. There were 127 exit interviews collected. These exit interviews served as the primary method by which offices within the academic success unit had collected data on reasons for student departure until the research site launched the student exit survey. Exit interviews were triggered when a student completed a leave form or were identified as transferring. Exit interviews were seemingly mandatory for student to complete but no enforcement mechanism was utilized. For this study, there were 117 interview notes, and each was connected to a student within the sample.

Data Analysis

To better understand the subclasses of nonreturning students at a private HEI, analysis was conducted by a team of researchers. This team consisted of academic faculty, staff, and both graduate/undergraduate assistants. Analysis was conducted in two phases. First, the population set of 1,091 student records was categorized into one subclass to fulfill the first purpose of the study. Directed content analysis (Hsieh & Shannon, 2005) guided the efforts of researchers by leveraging a conceptual framework posited by Hoyt and Winn (2004) for four subclasses of nonreturning students: drop-outs, stop-outs, opt-outs and transfer-outs. A fifth subclass of too early to determine was added because of the university business process of students being permitted two consecutive terms of unenrollment before a drop-out determination is made. The majority of student records (972 of 1,091) for this study did not contain university administered surveys, academic-advising notes, or exit interviews from which to determine intentionality of student departure. Even though considerably more data was collected and recorded by the institution, only 11% was applicable to understanding the intention of departing students—suggesting some efforts are less helpful to inform retention efforts. We elaborate more on this insight in our discussion and recommendations. Directed by Hoyt and Winn (2004) subclasses, researchers created data-sorting logic to deduce intentionality. Table 3 explains the logic applied to the population set. Please note that opt-out and drop-out are combined.

Table 3.
Subclass Logic for Directed Content Analysis.

Opt-out/drop-out Too early to determine

Students who were enrolled for one or more terms continuously, and their two most recent term statuses are not enrolled or cancelled. Opt-outs must include student open-ended responses to determine as opt-out-motivated departure Students who were enrolled for one or more terms continuously and only departed for a single term at the time of the study timeframe

Stop-out Transfer-out

Students who had noncontinuous enrollment or patterns of enrolled/not enrolled terms and/or not enrolled/enrolled terms, and they were either enrolled in spring 2021 or graduated by 2021 Students who enrolled at another institution for one or more terms

The research team was subdivided into two teams (i.e., Team A and Team B). Each subteam reviewed about half of the 1,091 student records. The student records included all of the available data discussed above. Using a strategy suggested by Hsieh and Shannon (2005), each student record was coded by one researcher into one of the five subclasses, and notes and highlights were made to support classification. For example, Researchers A1 and A2 reviewed the same 355 student records. If a disparity was identified between the two researchers in the coding of a primary subclass, a third researcher outside the subdivided team reviewed the record for final determination. Results were recorded in a leading qualitative data analysis computer software package, which aided in the development of an audit trail and interrater reliability (Lincoln & Guba, 1985).

To achieve the second purpose of the study, content analysis (Creswell & Creswell, 2018) was performed on the text of 119 exit interview field notes, university surveys, and advising notes. We began with start codes, formed categories, and finally identified emergent themes on the expressed departure intentions captured by the research site's business practices.

Finally, to achieve the third expressed purpose of the study, we used Fisher's exact test. We assessed the strength of the association between a student's characteristic group (i.e., sex, first-generation status, nonresident aliens, Pell grant recipients, transfer students, etc.) and their nonreturning subclass. Fisher's exact test was utilized to compare the proportion of certain characteristic groups with all other students included in the sample.

Positionality

Inherent in the design of action research, this research project relates to the professional action of its research team members. The seven-member research team comprised higher education professionals responsible for enrollment and retention, as well as student and academic support at the research site. The research team also included undergraduate and graduate students. Membership was predominately White, cis, and female.

Findings

Findings are presented in the order of the three purposes of the study. First, Table 4 presents the classifications of nonreturning students from the directed content analysis using Hoyt's and Winn's framework (2004). The largest subclass is transfer-out, representing 42% of the sample, with drop-outs representing 31%. Additionally, higher frequencies were observed in the fall cohorts from 2017 and generally decreasing to more recent cohorts in 2020.

Table 4.
Subclasses by Entry Cohort Term.

Cohort Drop-outs Stop-outs Opt-outs Transfer-outs Too early to determine

Fall

2017 131 63 0 134 12

2018 107 59 0 124 12

2019 59 51 1 136 23

2020 2 19 1 21 21

Spring

2018 18 7 0 15 3

2019 14 7 0 14 1

2020 10 3 0 18 5

Total 341 (31%) 209 (19%) 2 (1%) 462 (42%) 77 (7%)

The second purpose of the study was to summarize nonreturning students’ departing intentions between 2017 and 2020. Table 5 presents the four emergent themes that were identified from the content analysis (i.e., university experience and fit, financial, health, and academic), along with descriptions.

Table 5.
Theme Descriptions of Departing Intentions of Nonreturning Students.

Theme Start codes Theme description

University experience and fit Fit(general); fit(specific); inclusion; friendship; degree, major and career; identity; Greek life; faculty and staff; campus life; campus infrastructure; preplanned; athletic Around 100 of the participants described factors related to their perceived alignment with the university to be a primary factor in their decision to depart from the university. These factors were largely focused on perceived fit with their peers and a lack of peer community felt by some students outside of the Greek life program. Participants expressed frustrations with their peers’ lifestyles (e.g., entitlement, party habits, drinking) on campus, while others arrived at the research site with the intention to transfer later in their college experience. Finally, students described institutional degree program and major offerings as lacking their desired academic discipline.

Financial Cost of attendance, financial aid More than one-fifth of the nonreturning students expressed issues paying for and the increased burden of tuition, additional college expenses, and fees (e.g., social organization membership fees, housing), as well as overall attendance of the research site as their primary reason for departure. Some participants indicated financial aid policy or lack of financial aid as an obstacle to their continued enrollment.

Health Personal health, family health, pandemic-related About 20 of the participants expressed health concerns over the COVID-19 pandemic as related to their departure from the research site. Overall, this health factor ranged from health concerns related to being on campus, a desire to be closer to family, or a need to support their families. Health issues primarily related to personal well-being (e.g., mental, physical) as well as familial health were also used as factors in participant departure decisions.

Academic Academic challenge, degree constraints, academic credit policy Participants indicated academic concerns to be a factor for their departing status. This theme consisted of students expressing academic challenges due to the rigor of the university education at the research site. This theme was distinct from academic issues raised in previous themes; those were institutional offering-related, while this theme encompasses student academic experiences. Students named academic credit policy issues with transfer of community college credits and university curriculum requirements at the research site as well. Lastly, another issue that arose from the university curriculum was degree constraints caused by class requirements.

University Experience and Fit

The perceived fit between the participant and the university represented a significant reason for students choosing to leave the research site. As one student emphasized, “the university was not a good fit overall” for them. In addition to comments on general fit, other students expressed more specific issues with the university culture. For example, the “elitist and judgmental attitudes of my fellow students made me realize that [the research site] was not right for me.” Students expressed this lack of fit to be compounded for students outside of the Greek life program. “The entire social life of the school is dominated by Greek life,” remarked one student. Other participants decided to depart since realizing a lack of alignment between the university and their future career plans. This divergence led to departure because it was caused by a lack of specific degree or major. According to one, “for my degree and situation there is another that is a better fit.” Another expressed a need to “take time off to figure out what my passion is.” Some students went on to express concerns about the university faculty and staff as contributing to their intention to depart. For example, “school officials do not respond back to emails with alarming issues.” Others expressed concerns regarding the living requirement of the research site: “the requirement of living on campus … had a negative impact on my view of the university.” Campus infrastructure was also treated as an enabler for intentionality to leave. “The 20-min walks to class made finding the motivation to attend [class] difficult.” Athletic fit was also described by a handful of students, best recounted with the comment, “my style of play … [is] not a good fit for the program.” Lastly, to illustrate the theme of university fit, there were students who had the intention to leave from the beginning of their time at the research site; hence, their departure was preplanned. This sentiment was expressed through statements or comments like “I intended to leave [the research site] after 1 year.”

Financial

Not only did students intend to leave because of a lack of perceived fit with the university, but some students departed for financial reasons. Some participants felt “drowned by the cost of attendance.” Some students experienced financial stress, which was heightened by the site's financial aid policies. For example, one student felt discouraged upon realizing that “the financial aid decreases every time an outside scholarship has been earned.” Finally, changing aid packages during attendance was also cited as a contributing factor to departure.

Health

Thirdly, students conveyed health concerns to be another emergent theme for their intention to depart. While some participants left amid concerns for their personal health (e.g., mental and physical), others left to be caregivers to their family. To illustrate these concepts, one participant posited, “for someone with social anxiety … living on campus isn't easy.” Another health concern emerged in the data with the onset of the COVID-19 pandemic. Some participants wanted to “[live] closer to home … God forbid anything happen to my family.” Others preferred to depart their college education at the site because of concerns over safety: “I feel unsafe with the growing number of COVID-19 cases on campus,” a student enumerated in his exit survey.

Academic

Lastly, academic reasons captured the essence of some students’ intentions to leave. “I felt unchallenged or bored in many of my classes,” a student expressed while discussing the lack of academic rigor. On the other end of the dimension represented in this theme, some students felt overwhelmed by the university degree plan requirements. A student said, “I feasibly could not graduate in four years with my mechanical engineering degree with the horrid breadth/[curriculum] requirements.” Additionally, the site's academic credit policy was criticized for its restrictiveness: “You guys are wrong for not accepting most (if not all) community college credit hours,” a student noted. Therefore, academic reasons were a theme in student departure intentions. Table 6 presents the themes and their representations within each subclass. Overall, university fit was represented within 51 transfer-outs.

Table 6.
Theme Counts of Nonreturning Students by Subclass.

Theme Total Drop-out Stop-out Opt-out Transfer-out

University experience and fit 99 (63%) 0 4 7 88

Financial 34 (22%) 0 1 0 33

Health 17 (11%) 0 0 0 17

Academic 6 (4%) 0 0 0 6

Total 156 (100%) 0 4 7 144

Note. The too-early-to-determine subclass is not represented in the table above as no institutional data were available on the theme. Additionally, some participants might have expressed multiple departure intentions that could be counted in multiple themes.

The third purpose of the study was to compare nonreturning students by student characteristics as exported from the SIS. After performing a series of Fisher's exact tests, significant associations were uncovered between certain student characteristic groups and representation within the nonreturning population. Groups exhibiting greater-than-expected representation within the nonreturning population included nonresident aliens (p < .001), Pell grant recipients (p < .001), transfer students (p < .001), and first-generation students (p = .035). Alternatively, groups exhibiting less representation than anticipated within the nonretuning population included female students (p = .010), student athletes (p < .001), and students enrolled in a university-wide honors program (p < .001).

Discussion and Recommendations

Our findings yielded five insights that extend and corroborate the current literature: (a) drop-outs were the second largest subclass of nonreturning students, (b) opt-outs are significantly less represented within the nonreturning student population, (c) there is disproportionate representation of nonreturning students by some student characteristic groups, (d) an HEI's degree-seeking orientation may present a biased positionality affecting a student's well-being, and (e) intentionality is a derivative of the student voice.

First, of all the subclasses, drop-outs were the second-largest representative group (31%; n = 341) across the 1,091 cases encompassed in the study. Not surprising though, drop-out is often the singular term used to describe all nonretuning students in both the K–12 and HEI contexts and often does not reflect the nuances of all nonreturning students (Hoyt & Winn, 2004). Our study illustrates the need to provide nuance, though, as transfer-outs were the largest subclass represented at 42%. The emphasis and focus on drop-outs dominates current scholarly discussion and within the enrollment management literature (Brasher et al., 1980; Gathers, 1982; Tichenor, 1987). With our action-based research orientation, this often is a large group that is not well understood by HEIs. For instance, the research site's business practice is to allow students the ability to register for classes for two consecutive terms following a recent active term. If no action is taken by the student within two terms, they must reenroll before they are able to register for classes. By leaving the onus on the student, the institution misses an opportunity to collect the intention or to understand the reason for dropping out from the nonreturning student. For instance, institutions should reach out to students after one term of unenrollment to collect and record their intention to return. Nonreturning students encompassed in this subclass might indeed be better characterized by another lexical term. With additional institutional knowledge, the drop-out subclass may be further classified or better understood. Therefore, HEIs should develop mechanisms or a process to reach out to students who do not enroll for two consecutive terms to understand their unenrollment rather than merely assigning the drop-out subclassification.

Second, our study identified significantly fewer opt-outs (n = 2) as compared with any other subclass. This could be the result of students not leaving for this reason; however, it could also be that the university administered surveys and mechanism may not include questions that aid in the identification of opt-outs. To account for this possibility, HEIs should add questions that would better identify this subclass of nonreturning students. Examples of a simple yes/no question could include, “Are you opting out of a college degree in favor of another life opportunity/event?” This simple question could produce quicker classification by the institution of opt-outs. We suspect this subclass to be underrepresented merely by the limitations of the institutional information gathered from nonreturing students.

Third, our findings present a disproportionate representation of nonreturning students by some student characteristic groups (e.g., nonresident aliens, Pell grant-eligible students, and transfer students). Nonresident aliens make up 6% of the university's student population, yet account for 9% of nonreturning students. Fisher's exact test provided very strong evidence that nonresident aliens differ from all other students in their likelihood of not returning (p < .001). From this data, we can gather that this population of students may encounter more struggles with completing their degree than United States citizens. Furthermore, Pell grant recipients make up 13% of the university population, yet account for 17% of nonreturners, while transfer admits make up 15% of the university population, yet account for 20% of the nonreturning population. Similarly, statistical analysis from the Fisher's exact test provided evidence that these two characteristic groups differ from other students in terms of likelihood of not returning (p < .001 for both tests). These findings may imply that students who come from disadvantaged financial backgrounds may experience more trouble meeting financial obligations to maintain continuous enrollment. We can also reason that transfer students may require more resources and focus from the institution to ensure that they remain enrolled and graduate on time especially considering the challenges with transitioning to a new environment.

Fourth, through our research and within the related literature, there appears to be a biased positionality from which to assess a student's reason for departure. There is an assumption that graduation with a 4-year degree is the sole avenue to achieve success. This is problematic. For example, two student records within our study indicated a departure from the university to pursue a military career. While this non-degree-seeking objective may be contrary to the goals of student retention and graduation initiatives of HEIs, it may align with the general well-being or life goals of the student. Therefore, it is recommended that HEIs and future research diversify the definition of success in college. For instance, many universities are investing in entrepreneurial programs to cultivate business startups (Liguori et al., 2018). If a student opts out of college and starts a successful business, is this failure on behalf of the institution if it supported such endeavors with programs and infrastructure? HEIs ought to consider how they might include opt-outs in their success narratives alongside retention and graduation rates. In addition to successful businesses, opt-outs may include a desire to pursue military service, as was the case in our study for two participants.

Lastly, as researchers conducted coding of the sample (n = 1,091), it became clear that participant-articulated intentions played a key role in the categorization of subclasses (i.e., drop-out, stop-out, opt-out, and transfer-out). As discussed previously in the data analysis section, only 188 of the 1,091 student records contained university-administered surveys, academic-advising notes, or exit interviews—intention-focused data sources. This raised the question: Do student records that lack a clearly articulated intention constitute classification under the conceptual framework? Consulting Hoyt and Winn 2004 study in which students “were questioned about their reasons” for departure, we concluded that intentionality is a derivative of the student voice (p. 402). Two important recommendations extend from this conclusion: (a) In the case of the 542 records that did not contain clearly articulated intention-focused data (49.7%), we determined that using a subjective approach to sorting by subclass—using data not derived from the student voice (e.g., institutional records)—provided more value to the HEI than a broad label of “nonreturning student.” (b) It is important that HEI initiatives capture the student voice by questioning intentionality.

Limitations

While we believe our study to be sound, we note several limitations. First, when operationalizing nonreturning students and determining a subclass, we only made determinations on drop-out and stop-out categorizations after two terms had passed based upon university business practices. This strategy may have disproportionately skewed subclassifications in cohorts whose admission terms were in 2020 or 2021. Plus, we used National Clearing House Data which is only available for United States colleges and universities and may not have captured transfer-outs accurately for international students returning to their home country or pursuing school outside of the country. Second, this study was conducted in the summer of 2021 after almost 2 years of the COVID-19 pandemic. The global pandemic impacted all industries including higher education and may have contributed to uncharacteristically high number of stop-outs because students and families across the globe had dramatically changing personal and professional contexts within the timeframe. Further action research should be conducted to specially track COVID-19 cohorts of students or ensure that findings resonate in future years. Third, the 2018 student cohort was represented at a higher degree in the data because of the cohort's longer duration at the institution and greater timeframe to enter the status of nonreturning student. Fourth, prior to 2019, qualitative source data did not contain reasons for leaving the university (i.e., intentionality): the university administered survey and exit surveys were not instituted until 2019 at the charge of new provost-office administrator. Similarly, university forms documenting student reasons for departure (i.e., leave, withdraw, and cancel) were not widely circulated as part of a systematic process until 2019 when the academic success office was created.

Conclusion

Given the significant emphasis placed on student retention by higher education leaders and stakeholders alike, a greater understanding is needed on the subclasses of nonreturning students and their reasons for leaving their institutions. Especially in the wake of the COVID-19 pandemic, HEIs must consider the tools at their disposal to combat student attrition. Furthermore, our research emphasizes the disproportionate numbers of nonreturning students observed in certain student demographic groups. To combat this issue, HEIs should be motivated to provide additional resources to students who may be at higher risk for leaving their university. Perhaps most importantly, our findings demonstrate the importance of capturing the voice of nonreturning students through intention-focused questions. Aside from aiding in the classification of nonreturning students, having access to intention-focused data enables higher education leaders to better assess the drivers of student attrition. Thus, HEIs should maintain well-organized and consistent assessment tools that accurately identify the intentions of and give voice to nonreturning students.

Opt-out	Drop-out
Definition: Students who leave the institution having accomplished non-degree-seeking objectives Characteristics: Nondegree-seeking, avocational (i.e., college is hobby to main vocation); attend to earn certificates or some form of knowledge	Definition: Students who leave the institution permanently and do not reenroll toward degree-seeking objectives Characteristics: No plans to return; do not transfer to another HEI; usually motivated by financial reasons, family problems, or academic difficulties; tend to be older than typical students
Stop-out	Transfer-out
Definition: Students who leave the institution temporarily with plans to reenroll toward degree-seeking objectives Characteristics: Do not complete degree within the normal time schedule; health issues; financial reasons	Definition: Students who leave the institution permanently by transferring to another HEI toward degree-seeking objectives Characteristics: Status-driven; accessibility to different programs or program availability driven; financial reasons

Demographic	Count (n)	Percent (%)
Sex
Female	502	46%
Male	589	54%
Classification
1st-year	300	28%
2nd-year	575	53%
3rd-year	180	17%
4th-year	36	3%
Admission type
1st-year	803	74%
Transfer	216	20%
1st-year transfer	72	7%
Cohort term
Fall 2017	340	31%
Spring 2018	43	4%
Fall 2018	302	28%
Spring 2019	36	3%
Fall 2019	270	25%
Spring 2020	36	3%
Fall 2020	64	6%
Ethnicity
American Indian/Alaska Native	1	0%
Asian	80	7%
Black or African American	62	6%
Hispanic of any race	138	13%
Native Hawaiian/Other Pac-Islander	0	0%
Nonresident alien	97	9%
Race and ethnicity unknown	3	0%
Two or more races	57	5%
White	653	60%
Special populations
Pell-eligible in 1st year	188	17%
First-generation	63	6%
Athlete	3	0%
Fraternity and Sorority members	163	15%
University honors program	42	4%

Opt-out/drop-out	Too early to determine
Students who were enrolled for one or more terms continuously, and their two most recent term statuses are not enrolled or cancelled. Opt-outs must include student open-ended responses to determine as opt-out-motivated departure	Students who were enrolled for one or more terms continuously and only departed for a single term at the time of the study timeframe
Stop-out	Transfer-out
Students who had noncontinuous enrollment or patterns of enrolled/not enrolled terms and/or not enrolled/enrolled terms, and they were either enrolled in spring 2021 or graduated by 2021	Students who enrolled at another institution for one or more terms

Cohort	Drop-outs	Stop-outs	Opt-outs	Transfer-outs	Too early to determine
Fall
2017	131	63	0	134	12
2018	107	59	0	124	12
2019	59	51	1	136	23
2020	2	19	1	21	21
Spring
2018	18	7	0	15	3
2019	14	7	0	14	1
2020	10	3	0	18	5
Total	341 (31%)	209 (19%)	2 (1%)	462 (42%)	77 (7%)

Theme	Start codes	Theme description
University experience and fit	Fit(general); fit(specific); inclusion; friendship; degree, major and career; identity; Greek life; faculty and staff; campus life; campus infrastructure; preplanned; athletic	Around 100 of the participants described factors related to their perceived alignment with the university to be a primary factor in their decision to depart from the university. These factors were largely focused on perceived fit with their peers and a lack of peer community felt by some students outside of the Greek life program. Participants expressed frustrations with their peers’ lifestyles (e.g., entitlement, party habits, drinking) on campus, while others arrived at the research site with the intention to transfer later in their college experience. Finally, students described institutional degree program and major offerings as lacking their desired academic discipline.
Financial	Cost of attendance, financial aid	More than one-fifth of the nonreturning students expressed issues paying for and the increased burden of tuition, additional college expenses, and fees (e.g., social organization membership fees, housing), as well as overall attendance of the research site as their primary reason for departure. Some participants indicated financial aid policy or lack of financial aid as an obstacle to their continued enrollment.
Health	Personal health, family health, pandemic-related	About 20 of the participants expressed health concerns over the COVID-19 pandemic as related to their departure from the research site. Overall, this health factor ranged from health concerns related to being on campus, a desire to be closer to family, or a need to support their families. Health issues primarily related to personal well-being (e.g., mental, physical) as well as familial health were also used as factors in participant departure decisions.
Academic	Academic challenge, degree constraints, academic credit policy	Participants indicated academic concerns to be a factor for their departing status. This theme consisted of students expressing academic challenges due to the rigor of the university education at the research site. This theme was distinct from academic issues raised in previous themes; those were institutional offering-related, while this theme encompasses student academic experiences. Students named academic credit policy issues with transfer of community college credits and university curriculum requirements at the research site as well. Lastly, another issue that arose from the university curriculum was degree constraints caused by class requirements.

Theme	Total	Stop-out	Opt-out	Transfer-out
University experience and fit	99 (63%)	4	7	88
Financial	34 (22%)	1	0	33
Health	17 (11%)	0	0	17
Academic	6 (4%)	0	0	6
Total	156 (100%)	4	7	144

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.

ORCID iD

Dustin K. Grabsch

Lauren Sutro O'Brien

Dedeepya Chinnam

Zak Waddell

Author Biographies

Dustin K. Grabsch, PhD, is a proud first-generation college student and holds a bachelors, masters, and doctorate in leadershiprelated fields. Originally from Warren, Michigan, Dustin enjoys boating, hiking, DIY projects, and writing in his spare time. He currently serves students as the director of academic initiatives at Southern Methodist University. As the director, he oversees the Faculty in Residence program within the eleven Residential Commons; leads an interdisciplinary social science research team comprised of students, faculty and staff; facilitates a place-based community engagement program; and supports retention and graduation efforts within the co-curricular environment.

Lauren Sutro O'Brien has a B.A. and M.A. in Anthropology, and is currently a doctoral student in Education Leadership. Originally from Southern California, Lauren enjoys spending time with her family, hiking, and exploring new places. She currently serves undergraduates at Southern Methodist University as the Director of Student Success and Retention. As the director, she helps students navigate the barriers they face in college and achieve success.

Caroline Kirschner holds a bachelor's degree in business from Oklahoma State University and a master's degree in public affairs from the University of Texas at Austin. Caroline is a Texas native who enjoys reading, baking, and running. Caroline currently serves as a data analyst in the Office of Institutional Research at Southern Methodist University where she analyzes student data related to retention and graduation outcomes.

Dedeepya Chinnam is an undergraduate junior pursuing a Bachelor's in Business Analytics & Supply Chain Management, Statistics and Economics. Born and mostly raised in India, Dedeepya moved to the United States to pursue higher education. Personally, she loves reading books, playing sports, travelling, and spending time with her family. She currently serves in multiple roles on campus including being an undergraduate research assistant, a consultant to student organizations, student advisory board member and a member of student organizations executive boards. Her research interests include retention, leadership, motivation, organizational behavior, strategy and enterpreneurship.

Zak Waddell is a M.Ed. in Higher Education student at Southern Methodist University (SMU), holding a bachelor's degree from Texas A&M University (2002) and a Master of Humanities from the University of Dallas (2008). Zak was born and raised in Dallas, now spending most of his free time with two boys (2, 4), a wife and Pharaoh Hound. He enjoys soccer, cooking and karaoke. Zak currently serves as graduate assistant of academic initiatives at SMU. As graduate assistant, he collaborates on initiatives relating to assessment, research and planning.

Ryan Leibowitz is a graduate student pursuing a master's degree in applied statistics and data analytics. Originally from Frisco, Texas, he holds a B.S. in statistics from Southern Methodist University. When not in classes, he enjoys performing with the SMU marching band, reading, hiking, and traveling. Ryan currently serves as a graduate research assistant with the Office of Academic Initiatives at SMU. As a researcher, his interests include student retention, academic success, place-based community engagement, and mental health.

Michelle Madsen has her B.A. in Political Science and Public Relations and M.S. in Higher Education Administration. Originally from Madison, Wisconsin, Michelle enjoys anything outdoors including running, cycling, and hiking. She currently serves as an Associate Director for Residence Life and Student Housing at Southern Methodist University. As an Associate Director Michelle enjoys working with students, faculty, and staff in SMU's unique Residential Commons system, helping to make campus a home.

References

Adomssent

Godemann

Michelsen

(Eds.). (2007). ‘Sustainable university’ – holistic approach to sustainability in higher education institutions [special issue]. International Journal of Sustainability in Higher Education, 8(4), 385–402. https://doi.org/10.1108/14676370710823564.

Altrichter

Posch

. (1989). Does the ‘grounded theory’ approach offer a guiding paradigm for teacher research? Cambridge Journal of Education, 19(1), 21–31. https://doi.org/10.1080/0305764890190104

Amaral

L. P.

Martins

Gouveia

J. B

. (2015). Quest for a sustainable university: A review. International Journal of Sustainability in Higher Education, 16(2), 155–172. https://doi.org/10.1108/IJSHE-02-2013-0017

Brasher

D. E.

Jones

J. A.

Blom

D. L.

(1980). Attrition and retention. Report ED 224 367, U.S. Department of Education. https://files.eric.ed.gov/fulltext/ED224367.pdf

Britt

S. L.

Ammerman

D. A.

Barrett

S. F.

Jones

. (2017). Student loans, financial stress, and college student retention. Journal of Student Financial Aid, 47(1), 25–37. https://files.eric.ed.gov/fulltext/EJ1141137.pdf

Chowdhury

. (2006). Factors affecting retention of first-year Latino students in a private university. Doctoral dissertation, Morgan State University. ProQuest Dissertations Publishing.

Creswell

J. W.

Creswell

J. D.

(2018). Research design: Qualitative, quantitative, and mixed methods approaches (5th ed.). SAGE Publications.

Davis

Burgher

K. E.

Jefferson

. (2013). Predictive analytics for student retention: Programmatic evaluation and resource allocation. College and University, 89(2), 57–64. https://www.aacrao.org/docs/default-source/c-u-.pdfs/cuj8902_web.pdf?sfvrsn=63c95896_0

Davis

G. M.

Hanzsek-Brill

M. B.

Petzold

M. C.

Robinson

D. H

. (2019). Students’ sense of belonging: The development of a predictive retention model. Journal of the Scholarship of Teaching and Learning, 19(1), 117–127. https://doi.org/10.14434/josotl.v19i1.26787

10.

. (2011). Predicting student attrition with data mining methods. Journal of College Student Retention: Research, Theory & Practice, 13(1), 17–35. https://doi.org/10.2190/CS.13.1.b

11.

Demetriou

Powell

. (2014). Positive youth development and undergraduate student retention. Journal of College Student Retention: Research, Theory & Practice, 16(3), 419–444. https://doi.org/10.2190/CS.16.3.f

12.

Ewell

P. T.

(1984). Conducting student retention studies. Report ED 246 815. National Center for Higher Education Management Systems. https://files.eric.ed.gov/fulltext/ED246815.pdf

13.

Gal

. (2012). The role of mid-ranged universities in knowledge transfer in central and Eastern Europe - sustainable university strategies in the era of post-mass education. Journal of Economic Development, Environment and People, 1(1), 55–78. https://doi.org/10.1080/09654313.2011.586186

14.

Gathers

E. G.

(1982). A retention and attrition study of horizontal transfer, vertical transfer, and Native students at a selected university. Doctoral dissertation, Oklahoma State University. ProQuest Dissertations Publishing.

15.

Goodson

I. F.

Walker

(Eds.). (1991). Biography, identity and schooling: Episodes in educational research. Falmer Press.

16.

Haverila

M. J.

Haverila

McLaughlin

. (2020). Variables affecting the retention intentions of students in higher education institutions: A comparison between international and domestic students. Journal of International Students, 10(2), 358–382. https://doi.org/10.32674/jis.v10i2.1849

17.

Hermanowicz

J. C

. (2006). Reasons and reasoning for leaving college among the academic elite: Case study findings and implications. Journal of College Student Retention: Research, Theory & Practice, 8(1), 21–38. https://doi.org/10.2190/711H-UY47-CED1-8M47

18.

Hoyt

J. E.

Winn

B. A

. (2004). Understanding retention and college student bodies: Differences between drop-outs, stop-outs, opt-outs, and transfer-outs. NASPA Journal, 41(3), 395–417. https://doi.org/10.2202/1949-6605.1351

19.

Hsieh

H.-F.

Shannon

S. E

. (2005). Three approaches to qualitative content analysis. Qualitative Health Research, 15(9), 1277–1288. https://doi.org/10.1177/1049732305276687

20.

Ishitani

T. T.

Flood

L. D

. (2018). Student transfer-out behavior at four-year institutions. Research in Higher Education, 59(7), 825–846. https://doi.org/10.1007/s11162-017-9489-4

21.

James

J. L

. (2020). Students as stakeholders: Understanding expectations can increase student retention. Journal of College Student Retention: Research, Theory & Practice. Advance Online Publication. https://doi.org/10.1177/1521025119898844

22.

Keels

. (2013). Getting them enrolled is only half the battle: College success as a function of race or ethnicity, gender, and class. American Journal of Orthopsychiatry, 83(2–3), 310–322. https://doi.org/10.1111/ajop.12033

23.

Liguori

Winkler

Winkel

Marvel

M. R.

Keels

J. K.

van Gelderen

Noyes

. (2018). The entrepreneurship education imperative: Introducing EE&P. Entrepreneurship Education and Pedagogy, 1(1), 5–7. https://doi.org/10.1177/2515127417737290

24.

Lincoln

Y. S.

Guba

E. G.

(1985). Naturalistic inquiry. Sage Publications.

25.

Marsden

P. V

. (1990). Network data and measurement. Annual Review of Sociology, 16, 435–463. https://doi.org/10.1146/annurev.so.16.080190.002251

26.

McPherson

G. R.

(2016). Factors affecting student retention at a midsized private university. Doctoral dissertation, Ashland University. OhioLINK ETD Center. http://rave.ohiolink.edu/etdc/view?acc_num=ashland1465987353

27.

Melrose

M. J

. (2001). Maximizing the rigor of action research: Why would you want to? How could you? Field Methods, 13(2), 160–180. https://doi.org/10.1177/1525822X0101300203

28.

Millea

Wills

Elder

Molina

. (2018). What matters in college student success? Determinants of college retention and graduation rates. Education, 138(4), 309–322.

29.

National Student Clearning House. (2021). Clearinghouse academy. https://www.studentclearinghouse.org/academy/courses/introduction-to-enrollment-reporting/

30.

Neamţu

D. M.

Bejinaru

Hapenciuc

C. V.

Condratov

Stanciu

. (2020). Analysis and modelling of influence factors in the configuration of a sustainable university. Case study: ‘Ștefan Cel mare’ university of suceava. Amfiteatru Economic, 22(54), 391–410. https://doi.org/10.24818/EA/2020/54/391

31.

Patton

M. Q.

(1990). Qualitative evaluation and research methods (2nd ed.). SAGE Publications.

32.

Sagor

(2000). Guiding school improvement with action research. Association for Supervision and Curriculum Development.

33.

Shek

D. T. L.

Chai

. (2020). The impact of positive youth development attributes and life satisfaction on academic well-being: A longitudinal mediation study. Frontiers in Psychology, 11, Article 2126. https://doi.org/10.3389/fpsyg.2020.02126

34.

Soria

K. M.

Stebleton

M. J

. (2012). First-generation students’ academic engagement and retention. Teaching in Higher Education, 17(6), 673–685. https://doi.org/10.1080/13562517.2012.666735

35.

St. Amour

(2020, June 18). Stop-out programs expand amid pandemic. Inside Higher Ed. https://www.insidehighered.com/news/2020/06/18/pandemic-coincides-boom-projects-re-enroll-college-students

36.

Stephenson

A. L.

Yerger

D. B.

Heckert

D. A

. (2020). Recalibrating our understanding of retention. Journal of College Student Retention: Research, Theory & Practice, 22(1), 17–26. https://doi.org/10.1177/1521025117721852

37.

Taylor

R. D.

Oberle

Durlak

J. A.

Weissberg

R. P

. (2017). Promoting positive youth development through school-based social and emotional learning interventions: A meta-analysis of follow-up effects. Child Development, 88(4), 1156–1171. https://doi.org/10.1111/cdev.12864

38.

Thomas

(2016). Developing inclusive learning to improve the engagement, belonging, retention, and success of students from diverse groups. In Shah

Bennett

Southgate

(Eds.), Widening higher education participation: A global perspective (pp. 135–160). Chandos Publishing.

39.

Tichenor

(1987). St. Louis Community College nonreturning students: Retention problems or success stories? Report ED 287 551, U.S. Department of Education. https://files.eric.ed.gov/fulltext/ED287551

40.

Tinto

. (1975). Dropout from higher education: A theoretical synthesis of recent research. Review of educational research, 45(1), 89–125. https://doi.org/10.3102/00346543045001089

41.

Tinto

(1993). Leaving college: Rethinking the causes and cures of student attrition. The University of Chicago Press.

42.

Tinto

. (2006). Research and practice of student retention: What next? Journal of College Student Retention: Research, Theory & Practice, 8(1), 1–19. https://doi.org/10.2190/4YNU-4TMB-22DJ-AN4W

43.

Turner

M. A.

McDaniel

(2016, September). The fellowship initiative in context: Applying a positive youth development framework [brief]. Urban Institute. https://www.urban.org/sites/default/files/publication/84201/2000927-The-Fellowship-Initiative-in-Context-Applying-a-Positive-Youth-Development-Framework.pdf

44.

Webb

O. J.

Cotton

D. R. E

. (2018). Early withdrawal from higher education: A focus on academic experiences. Teaching in Higher Education, 23(7), 835–852. https://doi.org/10.1080/13562517.2018.1437130

45.

Zahid

Ghazali

Rahman

H. U

. (2017). A step towards sustainable university: A case of universiti teknologi PETRONAS (UTP) Malaysia. Global Business and Management Research, 9(1), 492–504.

46.

Zhao

C.-M.

Kuh

G. D

. (2004). Adding value: Learning communities and student engagement. Research in Higher Education, 45(2), 115–138. https://doi.org/10.1023/B:RIHE.0000015692.88534.de