Adult Learner Dropout in Online Education in the Post-Pandemic Era

Definition

Adult learner dropout is adults’ withdrawal or stop-out from formal or non-formal educational programs before successful completion. For adult learners, withdrawal often manifests as stop-out or temporary disengagement rather than permanent attrition, reflecting the episodic nature of their participation. Unlike traditional students, adult learners must often balance multiple life responsibilities—employment, caregiving, financial obligations, and community roles—while also pursuing education or training. Their vulnerability to attrition is further exacerbated by these overlapping demands, particularly when educational programs do not accommodate their situational and motivational needs. Adult learner dropout therefore requires a more dynamic understanding of persistence as a continuous negotiation between internal and external demands. Participation in online education has significantly expanded over the past two decades, particularly during and after the COVID-19 pandemic, as adult learners increasingly engage with digital platforms for work and communication. This exposure has enhanced their digital fluency, transforming their expectations and experiences of online learning. Thus, the underlying factors that influence adult learner dropout have also shifted—moving beyond technological and access-related barriers to instructional quality, engagement design, and relevance issues. In this evolving landscape, adult learner dropout can no longer be regarded as isolated or individual events. It is a systemic phenomenon emerging from dynamic interactions among psychological, pedagogical, contextual, and institutional factors.

1. Introduction

The rapid expansion of online education (OE) has transformed the landscape of adult learning, providing flexible and accessible opportunities across formal or non-formal contexts. With the use of digital technologies, adult learners can now engage in education regardless of geographical or temporal constraint, enabling lifelong learning fitting within complex personal and professional responsibilities [1,2]. The COVID-19 pandemic accelerated this transformation, and institutions, workplaces, and communities were compelled to adopt online delivery modes at an unprecedented scale [3,4].

This transformation has expanded participation in OE; however, persistent and emerging challenges remain. In spite of improved technological access, dropout rates among adult learners participating in OE continue to exceed those in traditional learning environments [5,6,7]. Recent research has found that, beyond technological access or digital literacy, pedagogical and engagement factors, such as the quality of instructional design, learner–instructor interaction, and the perceived relevance of course content, play a critical role in shaping persistence and attrition in online learning [8,9,10].

Existing dropout frameworks, dating to quite a bit before the pandemic, tend to be unable to account for the complexities of post-pandemic online learning environments. The widespread digitalization of work and education has transformed adult learners’ expectations, technological competencies, and motivational orientations. It has thus become increasingly important to reconceptualize dropout models by integrating instructional, contextual, and psychological dimensions to more accurately explain and prevent attrition in contemporary OE.

Building on this need for reconceptualization, this entry is guided by two central questions: (1) How have models and explanations of adult learner dropout evolved, particularly in the post-pandemic landscape? (2) How can a systemic, multi-level framework conceptualize dropout as an outcome of cross-level interactions rather than an individual deficit?

2. The Context of Online Education

OE—also known as distance education, web-based learning, or virtual learning—encompasses a wide range of instructional programs that are delivered through digital platforms [11]. Historically, the origins of OE can be traced to correspondence courses in the late nineteenth century. With the development of information and communication technologies and the proliferation of the internet, OE has expanded and now spans formal, non-formal, and informal learning, offering flexibility, accessibility, and scalability that go beyond traditional classroom instruction [12,13].

By the early 2000s, OE had become an essential component in global education systems. In the United States alone, in 2015, over six million learners were enrolled in at least one online or distance course, which represents nearly one-third of all enrolled college students [12]. More recently, data show that by Fall 2022, roughly 54% of US college students had taken at least one online class, which signaled a novel normal plateau for online enrolment [14]. At a global level, organizations such as the United Nations Educational, Scientific and Cultural Organization (UNESCO) and the Organisation for Economic Co-operation and Development (OECD) continue to emphasize OE as a mechanism for the democratization of access and for supporting workforce reskilling and lifelong learning, particularly in regions having limited educational infrastructure [15,16].

The COVID-19 pandemic dramatically accelerated these trends. In a span of months, worldwide institutions transitioned to online delivery, leading to a rapid and unprecedented expansion of virtual learning across systems of higher education [17,18]. In Europe, 92% of institutions reported providing remote instruction during the pandemic [19], marking the normalization of online learning. Thus, OE has evolved into a core pillar for global education, being applied in formal schooling, continuing education, and community-based learning initiatives.

Despite the widespread expansion of OE, particularly in adults, persistent challenges remain, in particular the high dropout rates and low learner persistence in OE, both of which continue to threaten institutional sustainability. Recent evidence highlights that online learners already experience substantially lower retention rates—with 10% to 20% lower persistence than students in traditional face-to-face courses and dropout rates ranging from 40% to 80% [20]. This concern is amplified for adult learners: students entering college at age 21 or older persist at rates more than 30 percentage points lower than those aged 20 or younger, underscoring the pronounced challenges faced by adults in sustaining continuous participation in OE [21].

While early research primarily attributed these disparities to issues of technological access and digital readiness, recent studies have indicated a paradigmatic shift toward pedagogical, motivational, and design-related determinants of persistence [7,8,10]. In this evolving context, the influence of technological accessibility has diminished within the post–COVID-19 era, as widespread digital exposure has made access less decisive in learners’ sustained engagement and success. Instead, the quality of instructional design, the perceived relevance of the learning experience, and the degree of cognitive and affective engagement are more decisive predictors for persistence in online adult learning [7,8]. Addressing these complexities requires the application of integrative frameworks that foreground the interplay between instructional quality, learner engagement, and continuous institutional support, across the learning process.

3. Adult Learners: Characteristics and Participation in Online Education

Adult learners are unlike traditional students in both their motivation and their context. They are typically self-directed, goal-oriented, and experience-based, in search of education that is relevant to their professional and personal lives [22,23]. Often, these learners must balance their studies with multiple responsibilities beyond education, making flexibility and relevance key to sustaining their engagement [1].

OE aligns well with these learners’ needs, providing asynchronous, flexible opportunities for continuous learning. Many adults take courses online for career advancement, upskilling, or requalification [6,16]. According to the OECD [16], approximately 45% of adults in member countries are participants in formal or non-formal learning activities, and these are increasingly being integrated with digital and hybrid formats. In Europe, systems of adult and continuing education have likewise accelerated the adoption of online and blended modalities beginning with the COVID-19 pandemic, reflecting a long-term digital transformation of lifelong learning [16,17,19].

However, the same features that attract adults to OE can also introduce unique challenges. Technological barriers, including limited computer self-efficacy and digital literacy, hinder learner participation [24]. Only 55% of European adults aged 16–74 possess basic digital skills, and this shows sharp disparities by education level [25]. Furthermore, environmental pressures—long working hours, caregiving, and financial obligations—can reduce persistence in education [5,26].

Taken together, these barriers highlight the dual reality of adult learning in the digital era: OE expands access, but it does not guarantee equitable or sustained participation. In addition, in the post-pandemic context, participation and persistence are no longer solely constrained by technological readiness but are deeply intertwined with pedagogical alignment, motivational support, and institutional responsiveness—factors that are explored in the following sections.

4. Evolution of Models Explaining Adult Learners’ Dropout from Online Education

Early theories of student attrition, such as Tinto’s Student Integration Model [27] and Bean and Metzner’s Nontraditional Student Attrition Model [28], provided foundational insights into students’ withdrawal from academic programs. Tinto’s model [27] emphasized academic and social integration, that is, the degree to which students feel intellectually and socially connected to their institution, taking it as central to persistence. Bean and Metzner [28] extended this framework, treating of nontraditional students, highlighting the additional role of environmental variables such as finances, employment, and family obligations. However, both models were developed in campus-based settings, which assumed that learners could devote substantial time and resources to study. They largely overlooked situational constraints such as employment, caregiving, and time scarcity, which fundamentally define the adult learning experience, in particular in online environments.

A significant conceptual turning point occurred with Rovai’s Composite Persistence Model [29], which synthesized insights from Tinto [27], Bean and Metzner [28], and adult learning theories. Rovai [29] was among the first to explicitly adapt persistence theory to OE, integrating pre-admission variables (e.g., student characteristics, computer skills, and self-directedness) and post-admission variables (e.g., satisfaction, social integration, and external support). His model emphasized that adult learners’ continuation decisions are not shaped only by academic integration but also by external responsibilities, such as work and family commitments. Rovai’s approach exhibited the early recognition that persistence in online learning arises from the interaction among individual readiness, course design, and life context. Importantly, he identified student–instructor and student–student interactions as key mechanisms of belonging, foreshadowing later research on engagement’s qualitative dimensions.

Building on Rovai’s conceptual foundation, Park [30] and Park and Choi [6] further refined the persistence framework, reflecting the realities of adult learners in online non-degree training programs. Their work embodied a decisive shift from theoretical models toward empirically validated explanations of dropout in workplace-oriented e-learning contexts. Park [30] categorized the determinants of dropout into four domains: learner characteristics (e.g., age, gender, prior education), prior skills (e.g., computer literacy and time management), external factors (e.g., family and employer support, workload, and life responsibilities), and internal factors (e.g., motivation, satisfaction, and the relevance of the course).

Using regression analyses of adult e-learners, Park and Choi [6] demonstrated that external support, particularly that from family and employers, and the perceived relevance of the course content were the most significant predictors for persistence, but demographic factors such as age and gender exerted minimal influence. Their findings indicate that in adult learners, persistence decisions are shaped primarily by alignment between instructional content and professional or personal goals. Learners who identify strong relevance between course materials and their work or life contexts were substantially more likely to complete their studies, while a lack of perceived relevance and external support increased the likelihood of withdrawal.

Crucially, Park and Choi [6] advanced the understanding of online adult persistence, showing that motivation and satisfaction mediate the relationship between instructional design and persistence. This insight reframed dropout theory, taking it beyond structural explanations that were centered on access and demographics and toward pedagogical and motivational perspectives that were rooted in adults’ lived experiences of online learning. In this way, their study provided a pivotal empirical bridge extending between traditional persistence models and emerging scholarship on online adult learning.

Most prior frameworks have either been conceptual or limited to short-term, non-degree training. Addressing this gap, Choi and Kim [31] empirically validated a data-driven model for use in the study of adult learners in online degree programs, refining the prototype proposed by Choi [32]. Using administrative data from 3462 students across 15 cyber-university programs, their analysis demonstrated that persistence is shaped by the interaction of learner, instructional, and contextual factors.

The model identified scholastic aptitude and instrumental studying motive (e.g., career advancement or earning credentials) as key pre-admission predictors for persistence, and external physical constraints—work, family, or financial burdens—were the strongest barriers. For the internal factors, interaction quality, motivation, and academic achievement (GPA) emerged as critical drivers for persistence, and they highlighted a feedback loop between motivation and performance. Collectively, their findings reframed dropout as other than personal failure, representing a systemic misalignment between program designs and learners’ contexts.

This trajectory, moving from Tinto’s integration theory [27] to the models that were proposed by Rovai [29], Park and Choi [6], and Choi and Kim [31], shows an evolution of adult learner dropout research, moving from static, demographic explanations toward multifactorial, empirically grounded perspectives. These earlier models formed the groundwork for understanding the interplay of learner characteristics, instructional design, and external circumstances. The rapid transformation of OE during and after the COVID-19 pandemic reshaped the research landscape, exposing the limitations of pre-pandemic dropout frameworks. Earlier models—focused on academic integration, environmental constraints, or individual motivation—treated these determinants largely in isolation. Recent studies, however, show that instructional design, affective engagement, digital resilience, and institutional adaptability interact in ways that earlier frameworks did not account for. These shifts underscore the need for a systemic, multi-level model capable of explaining persistence under contemporary conditions.

Recent meta-analytic evidence shows that instructional and pedagogical factors presently outweigh purely technical considerations in the shaping of persistence in online adult education. A systematic review by Rahmani et al. [7] analyzed more than a decade of research on online higher education, finding that among the wide range of influencing factors, pedagogical quality, learner motivation, and institutional support emerged more frequently and prominently as contributors to drop out than purely demographic or technological variables. In particular, improvements in instructional design, feedback mechanisms, and learner-support systems emerged as crucial interventions for sustaining persistence. Complementary studies of online learning and instructional design indicate that teaching presence, course relevance, and interaction quality are key determinants for learner engagement and persistence [9,29,33]. Effective instructional design, which is characterized by empathy, usability, and authenticity, transforms digital platforms from mere delivery tools into dynamic learning ecosystems. Consequently, scholarship currently calls for integrative instructional models to connect design quality, affective support, and adaptive feedback, instead of focusing solely on infrastructure [17].

These findings represent a paradigm shift in the study of online adult education: transitioning away from a focus on technological access and individual attributes and toward a systemic understanding of how instructional, motivational, and institutional dynamics shape learner persistence. After the pandemic era, the emphasis has moved beyond descriptive analyses of isolated variables toward predictive, feedback-based approaches modeling the interdependence of multiple levels of systems. Accordingly, the following section introduces the Systemic–Interactional Predictive Model of Adult Learner Dropout, which synthesizes these insights into a multilayered, data-informed framework that captures the dynamic interconnections among psychological, pedagogical, contextual, institutional, and analytic domains that collectively sustain—or disrupt—persistence in online adult education.

5. Major Variables Affecting Adult Learners’ Dropout from Online Education

Dropout of adult learners from OE is best understood as a systemic and interactional phenomenon that emerges from the dynamic interplay of multiple interdependent domains. The Systemic–Interactional Predictive Model that is proposed in this study conceptualizes dropout as an emergent outcome of cross-level misalignment in a continuously adaptive ecosystem (see Figure 1.). Each level—from the micro-level of learner cognition and motivation to the meta one of functioning as the adaptive intelligence of a system that integrates predictive analytics and feedback across all levels—exerts a distinct influence, but all operate in an integrated regulatory network co-regulating persistence. The following sections elaborate on five interrelated levels and their respective contributions to understanding dropout and persistence in the post-pandemic learning landscape.

5.1. Micro-Level: Psychological and Cognitive Domain

Persistence begins with the learner. At the micro-level, motivation, self-regulation, metacognition, and digital resilience underpin adults’ capacity to sustain learning while managing work and family responsibilities. Foundational theories emphasize goal orientation, autonomy, competence, and reflective monitoring as central to adult learning [1,34,35], and empirical research confirms that metacognitive regulation and learning motivation jointly predict persistence in online environments [31]. Post-pandemic scholarship reframes digital competence as digital resilience, highlighting learners’ ability to manage technological uncertainty, overload, and interruptions while maintaining continuity in online settings [36,37,38]. This micro-level adaptability remains a foundational determinant of persistence in contemporary OE.

5.2. Meso-Level: Pedagogical and Engagement Domain

At the meso-level, instructional design and engagement structures shape how learners translate individual readiness into active participation. Pedagogical coherence—alignment among objectives, activities, and assessments—strongly influences persistence [7,8,9,33], while unclear structure or excessive cognitive load can increase dropout intentions [6]. Recent learning analytics studies identify patterns such as meaningful interaction quality, navigation behaviors, and timely feedback as reliable indicators of sustained engagement [31,39,40,41]. As a mediating hub, the meso-level channels learner motivation upward to institutional systems while receiving adaptive adjustments guided by analytics and instructional design.

5.3. Exo-Level: Socio-Environmental Resilience Domain

The exo-level captures the social, occupational, and environmental contexts that frame adult learning. External pressures—including workload, caregiving, and financial strain—remain major barriers to persistence [6,42,43]. Although OE provides flexibility, inequities in digital access, study space, and family support continue to shape participation. Institutions increasingly respond with modular curricula, micro-credentials, and asynchronous formats that enable adults to pause and re-enter learning [31,44]. Supportive supervisors, family networks, and equitable resource access strengthen resilience by buffering the external disruptions that adult learners frequently encounter.

5.4. Macro-Level: Institutional Adaptability Domain

At the macro-level, institutional policies and cultures determine the extent to which OE environments can support diverse adult learners. The pandemic accelerated data-informed decision-making, flexible scheduling, and expanded well-being infrastructures [7,10]. Institutional adaptability involves redesigning curricula and assessment policies in response to learner data, while cultivating empathy, inclusion, and lifelong learning values. Rigid administrative requirements can heighten attrition, whereas early-alert systems, advising, and integrated support services foster persistence by enabling institutions to realign structures with learner needs [7,9,17].

5.5. Meta-Level: Predictive and Adaptive Feedback Domain

The meta-level integrates behavioral and contextual information from across the system through analytics and AI-based feedback mechanisms [44,45]. These tools detect early disengagement—e.g., reduced logins, incomplete tasks, weakened interaction networks—and coordinate timely intervention across micro, meso, and macro levels [7]. Ethical use of analytics requires transparency, privacy protections, and respect for learner autonomy [46,47]. From a systemic–interactional perspective, analytics transform dropout from an individual problem into a system-level signal that activates coordinated responses to restore balance across the learning ecosystem.

5.6. Cross-Level Dynamics

Dropout rarely results from a single factor; rather, it emerges from misalignment across system levels. Strong motivation at the micro-level cannot compensate for incoherent course structures at the meso-level or inflexible institutional policies at the macro-level. Conversely, coordinated analytics and adaptive feedback at the meta-level can strengthen persistence by aligning learner experience, instructional design, external supports, and institutional structures. Post-pandemic evidence underscores that persistence depends on dynamic equilibrium across levels, and dropout reflects systemic imbalance rather than individual inadequacy.

6. Implications and Future Directions

The evolving nature of adult learners and the post-pandemic transformation of OE require systemic reinterpretation of frameworks of persistence. Earlier models of dropout—though foundational—reflect conditions developed in contexts no longer in alignment with today’s data-rich, flexible and institutionally fluid learning environments. The Systemic–Interactional Predictive Model reconceptualizes persistence as an adaptive equilibrium maintained in continuous feedback among psychological, pedagogical, contextual, institutional, and analytic systems. Thus, dropout represents not an endpoint but a systemic signal of imbalance that can be mitigated through coordinated, data-informed responsiveness.

Traditional theories [6,27,28,29] have explained dropout with linear causality, but post-pandemic evidence reveals nonlinear interactions among motivation, flexibility, instructional design, and data feedback. Recent works [7,44,45] have emphasized multilevel analytics and AI-based dashboards that connect micro-level behavior with institutional response. However, such data-driven systems must remain ethical and human-centered, ensuring transparency, privacy, and learner agency [46,47].

Instructional design mediates across levels, aligning learner readiness with institutional adaptability in continuous feedback. The use of empathy, usability, and authenticity as design principles enables real-time responsiveness to learner data [31,33]. Future research should examine how design and analytics coevolve in sustaining coherence and engagement.

The pandemic blurred the boundaries between work, family, and education, intensifying adults’ emotional and cognitive burdens [7]. For the Systemic–Interactional Predictive Model, psychological sustainability is cross-level regulatory process where sustained motivation and affective balance emerge from the interplay between individual self-regulation and systemic support mechanisms.

Learners’ emotional resilience is reinforced in multilayered feedback loops: involving self-regulation at the micro-level, empathetic and inclusive pedagogy at the meso-level, and a responsive institutional climate at the macro-level [1,9]. Integrated services focusing on well-being and workload balance strengthen affective persistence [10,17].

Sustainable persistence relies on institutional and policy alignment. Early-warning analytics and flexible participation mechanisms (such as modular curricula and micro-credentials) support episodic engagement [33,43], while inclusive lifelong learning frameworks ensure equity [2,15]. Increased coordination between institutional responsiveness and predictive intelligence [7] can transform dropout prevention into proactive participation enablement.

The ongoing transformation of OE underscores that dropout functions as a systemic feedback signal—a disruption in coordination among learner experience, instructional design, and institutional adaptability. In this context, persistence emerges as the restoration of an adaptive equilibrium through responsive, data-informed feedback. Consequently, refined frameworks must adopt systems thinking, embedding feedback, empathy, and responsiveness at every level. Through such alignment, OE can move from merely preventing dropout to actively cultivating persistence, lifelong engagement, adaptability, and well-being.

Although the model provides a comprehensive conceptual structure, it remains theoretical and has not yet been empirically validated across diverse adult learner populations or institutional contexts. Future research should examine how cross-level interactions unfold in practice and test predictive indicators of persistence within various OE environments. Moreover, the model does not fully incorporate structural inequalities—such as gendered caregiving responsibilities, socioeconomic disparities, and disability-related accessibility barriers—that may shape persistence across multiple levels. Understanding how these intersecting inequalities influence system alignment represents an important direction for further research. Comparative studies across regions, program types, and institutional structures would further clarify how contextual variation affects system-level coordination and the conditions under which adult learners persist or withdraw.