Until somewhat recently, student persistence had been treated more like an afterthought than a priority in higher ed, but the emerging practice of dedication to persistence and student retention is now growing into a major institutional force. As such, higher ed leaders are focusing more intently on what it takes to keep students effective, engaged, and – ultimately – enrolled. What has been born from this effort has since gone on to help influence institutional approaches to persistence.

Now, with a changing student body and additional disruption in the realm of course delivery, student persistence models have forked. It is no longer enough to build a bright and lively campus with dedicated program support centers. Modern students need to connect their educational experience with their career, engage in real-world practice, and work according to their own schedule. Students who find their needs unmet are also finding it easier to transfer or drop out.

As the student population modernizes and changes, so too must the approaches to persistence. This is not to say that traditional models have gone sour; there are still plenty of uses for those models where traditional students are concerned. What’s especially relevant to the modern landscape is the fork. Schools that can offer both will have the greatest advantage and will be able to keep retain their students up to graduation and potentially keep them engaged beyond that.

What follows, then, is a review of traditional and contemporary student persistence models. Following that, a list of actionable strategies that may help retain face-to-face and online students at both the bachelor’s and graduate level.

Traditional Persistence Models

The Undergraduate Dropout Process Model

One of the first major models that emerged from the early days of persistence research came from William G. Spady’s Dropouts from Higher Education: An Interdisciplinary Review and Synthesis. In it, he surmised that institutions had a large role to play in student persistence, and that rather than just one, there were two variables, or systems, that could affect whether a student would stay at the institution: academic and social. Moreover, Spady noted that there were at least two factors in each of those systems that influence a student. Where academic factors were concerned, Spady delineated them into grades and intellectual development; and as for factors to support the social system, he identified friendship and normative congruence. The factors he identified have since been expanded, contorted, stretched, and changed a number of times, but his largest influence was in giving the institution responsibility for persistence.

Figure 1: Spady’s Undergraduate Dropout Process Model (Aljohani 2016)

Later, Spady’s model would serve as a launching point for other educational theorists, namely Tinto’s Model of Institutional Departure (1975, 1993), and Bean’s Student Attrition Model (1980, 1982).

The Institutional Departure Model

Vincent Tinto’s Institutional Departure Model, which was drafted in 1975 but did not reach its final version until 1993, builds on Spady’s, but ultimately asks a little bit more of the student as they build a relationship with the school. For Tinto, the social aspect of persistence was demarcated by the student’s ability to interact with the social and academic systems at the institution.

Figure 2: Tinto’s Institutional Departure Model (Aljohani (2016)

What Tinto realized, is students bring associations and expectations with them in their first year. He mapped out a process that begins with the student’s prior associations, but allows for those to be weakened or strengthened based on the way the student is incorporated into the institutional community. Successful incorporation might find those goals changed by the time the student has shed connections to old communities in lieu of their new community. In the case that student associates and expectations are less malleable, students may find themselves at a higher risk of dropping out.

The Student Attrition Model

In 1980, John P. Bean put forth his own model for student persistence that he called the Synthesis of a Theoretical Model of Student Attrition, or more succinctly, the Student Attrition Model. Here, he too built on the work of his predecessors. He differentiated by arguing that the motivations for departure are similar to those seen in an employee unsatisfied with their career or employer. Where obvious differences occurred between student and employee, such as pay, Bean found it useful to substitute equitable variables from the collegiate experience, such as GPA, student development, and career relevance. In a revised version of this model, Bean would include a set of four instrumental variables with his theory: background, organizational, environmental, and attitudinal and outcome variables. What he ultimately found was that these institutional factors played the largest role in student persistence, and that by remixing the variables, his model could be made to apply to nearly any industry.

Figure 3: The Student Attrition Model (Aljohani 2016)

Contemporary/Online Student Persistence Models

The Nontraditional Undergraduate Student Attrition Model

Shortly after publishing the Student Attrition Model, Bean revisited student persistence with help from Barbara Metzner and developed a model for nontraditional students, and more specifically, commuter students. What they understood was that the institutional integration and culture-building importance is not elevated to the same importance as it would be with traditional students. For the commuter student, environmental and external factors were the main forces acting against persistence.

Like Bean’s previous model, the Nontraditional Model makes room for four sets of variables: academic performance, intent to leave, background, and environmental factors.

Figure 4: The Nontraditional Undergraduate Student Attrition Model (Aljohani 2016)

The Self-Determination Theory of Student Persistence

Using student motivation as the research driver, Kuan-Chung Chen and Syh-Jong Jang applied the theory of self-determination to persistence in online and distance learning. The underpinning of the self-determination theory — which has found its way into politics, religion, healthcare, and similarly influential parts of society — is that individuals have three basic needs: autonomy, competency, and relatedness. When these needs are satisfied, When these needs are satisfied, we experience a heightened sense of self and increased protential for personal growth. It is not hard, then, to see how this might be adapted to online persistence.

The three pillars of self-determination can be applied to the elements of distance education, such as flexible learning, computer-mediated communication and social interaction, and technical proficiency. What Chen and Jang eventually conclude is that supporting an online student’s three basic needs positively affected their self-determination. Additionally, online persistence can be influenced by course factors and support services.

Figure 5: The Hypothesized Model of Self-Determination Theory (Chen & Jang 2010)

Factors that Impact Student Persistence

Each of the previous models uses specific factors to measure persistence among students. Factors might be broadly categorized as social, academic, or background. Each of these factors can be more narrowly defined by carving out individual influences. In the case of Spady’s model, the academic category is broken into influences such as grades and intellectual development (even now, Spady’s initial classifications are consistent with the influences seen in contemporary models, albeit somewhat unremarkably, given their broad nature). What has emerged recently are new technologies and modalities that disrupt the theoretical models constructed by Spady, Tinto, and Bean. Students have different expectations of their education as well as different financial and social concerns relevant to their schooling.

The charts that follow deliver a high-level view of the most common persistence categories and influencers from the major models.

Face-to-Face Students

The face-to-face metric includes academic, social, circumstantial, and personal influences. Initial thought relied on academic and social influences, whereas more recent developments have brought certain circumstantial and personal influences into the equation.

GradesNormative CongruenceFinancialMotivation
Intellectual GrowthFriendshipEthnicitySelf-Perception
Uneven SkillsInstitutional FitGeographicBehaviors
Academic PlanInstitutional SupportDemographicProblem-Solving Skill


The genesis of the academic model lies in Spady’s initial research but has widened to include uneven skills and academic plan. Sternberg’s 2013 research surmised that students who enter college without a rounded education (especially in STEM areas) face an uphill climb from day 1. Part of the issue is a reliance on standard aptitude tests, which he notes only serve as accurate forecasting agents 25 percent of the time. The knowledge needed for success at a particular institution, therefore, is not known.

Similarly, the lack of an academic plan contributes to a feeling of academic displacement for many students. This is made most clear with students who fail to recognize the connection between their coursework and a future career. In the instances when such a plan is formalized, students have been significantly more likely to persist.

As for grades and intellectual growth, Spady’s initial findings ring true: A student who achieves high marks and experiences intellectual growth is more likely to continue.


Where face-to-face students are concerned — and in the traditional campus setting — there is significant data to suggest that social factors are a leading influence on student persistence. Spady’s idea of normative congruence fits well with more recent research into how friendships and student communities help learners locate their place within the institution.

Institutional support itself comes in two forms. The first regards the relationship between students and their faculty. In many cases — especially in graduate settings or with freshmen — this relationship has a key role in supporting student persistence. Creating these types of institutional bonds should be a primary goal for schools concerned about attrition. The second comes from the way schools support student goals and lifestyles. Those could come in the form of dedicated learning centers and extracurricular clubs. Both are instrumental in forging a sense of community and addressing the need for students to feel supported.

As for institutional fit, research suggests that when accurate student profiles can be built (most likely by the admissions staff and director of admissions), institutions can better support the needs of the student and promote persistence.


Unsurprisingly, circumstantial factors influence student retention quite a bit. With financial factors leading the pack, challenging personal circumstances can place significant strain and additional stress on students. What is most discouraging is that in the face of financial difficulty, penny-wise students may make the decision to dropout — even when taking on debt and persisting would likely be a pathway toward long-term financial improvement. Fortunately, many colleges are finding creative ways to alleviate some of the financial burden by offering online or alternative pathways to a degree.

Background information such as ethnicity, geographic, and socioeconomic indicators continue to show ties to student persistence. Bean used background as a category to wrangle these factors together but made no mention of financial constraints. His factors have been broken up here to sharpen the obliqueness of this category.


While it is true that more emphasis has been placed on these indicators in recent years, their presence has been acknowledged since the early days of persistence research. Certainly in Spady’s as well as Tinto’s early models, it is evident to see some of the personal concerns overlapping into social concerns. By carving out a new category for these concerns, though, schools can help address them in a more creative and effective manner.