Unravelling the Mechanisms of Student Negative Emotions in Technology-Mediated Intermittent Dropout: A Mixed-Methods fsQCA Study

Unravelling the Mechanisms of Student Negative Emotions in Technology-Mediated Intermittent Dropout: A Mixed-Methods fsQCA Study | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Unravelling the Mechanisms of Student Negative Emotions in Technology-Mediated Intermittent Dropout: A Mixed-Methods fsQCA Study Yao Yao, Mingda Wang, Yang Zhang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6971678/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 8 You are reading this latest preprint version Abstract This study set out to unravel the complex mechanisms by which student negative emotions contribute to intermittent dropout in technology-mediated learning environments. By using a mixed-methods approach – qualitative interviews followed by fuzzy-set qualitative comparative analysis – we were able to identify distinct causal pathways leading to dropout, rather than attributing it to any single factor. The findings demonstrate that dropout is a multi-faceted phenomenon : it can result from an overload of work and lack of support provoking anxiety, from social isolation breeding apathy, from personal lapses in coping allowing frustration to fester, or often a combination of these elements. In all cases, negative emotions are at the core of the story , serving as the immediate precursors to a student's decision to disengage. This underscores a key point for both theory and practice: emotional experiences are not ancillary to academic outcomes, but fundamentally intertwined with them . student negative emotion Intermittent Dropout education technology application mechanism fsQCA Figures Figure 1 1 Introduction By advocating with the benefits of on-line and technology assisted learning environments, they have expanded access to education, however they also face the challenges of student disengagement and dropout during the process. Unlike face to face teaching context, the online courses often report higher attrition rates. In a survey conducted with 300,000 students, it revealed that over 70% of them felt disengaged during the first learning period (Li 2022 ). The intermittent dropout, where students are temporarily disengaged or “drop out” during the virtual course learning, undermines the students' ability of consistent learning and engagement and raise the concern for educators and related stakeholders (Rahmani, Groot, and Rahmani 2024). Antecedent literature has addressed the negative emotion affection as a pivotal role for students dropout behaviour. In contrast with face to face teaching, teachers always have the opportunities to observe student dropout behaviour with reminding and feedback to ease frustration or confusion immediately, online learners have to struggle with negative emotion in isolation (Eberle and Hobrecht 2021). Especially during the COVID-19 pandemic, studies claimed the high level anxiety, stress with other negative emotions among online environment learning students, which lead to excessive challenges mentally and chances of drop out (Baltà-Salvador et al. 2021 ; Li 2022 ). Based on the previous literature, the learners with high negative learning emotions – anxiety, boredom, or depression- have more intention of withdraw or dropout from study comparing contradictory (Bekker, Rothmann, and Kloppers 2023; Respondek et al. 2017 ; Tze, Daniels, and Klassen 2016). However, studies on these emotional factors integration are still scant on the models of online persistence, while the mainstream interests are perspectives from cognitive, demographic, or infrastructural factors(Delahunty, Verenikina, and Jones 2014; Sanborn 2022 ). Various theoretical frameworks have been explored the mechanisms of emotional experiences might lead to learners dropout in digital assistant learning context. The Emotional Regulation Theory (ERT) stated mechanism of an individual manage and respond to their emotional states (Andrei Sitar-Tăut, Mican, and Ioan Moisescu 2024; Lv et al. 2024 ; McRae 2016 ). As for learning context, the learner emotional regulation ability could affect its disengagement possibility (Boekaerts 2011 ; Liu and Liu 2025 ; Rahmani et al. 2024 ). Studies have revealed that cognitive reappraisal could eliminate anxiety or depression level or strategy of suppression expression may lead to worsen stress (Preston et al. 2022 ; Xu et al. 2020 ). Actually, individual relying on suppression emotion experienced worst by negative factors and tend to report their feeling of more lonely, which in turn can heighten anxiety (Eres et al. 2021 ). This denotes that students cannot manage depression or anxiety successfully in online learning context, which may lead to burnout even learning dropout (Li and Yang 2025; Peng et al. 2023 ). Another relevant lens is the concept of information overload in digital environments. Online course are conducted with dense reading materials, multi-media visual and audio learning files, discussion activities, and countless notifications (Roopashree et al. 2024 ). Referring to Cognitive Overload Theory (COT), people attention is finite while overwhelming content and multi-dimensional requirements are likely trigger fatigue with distraction (Koundal et al. 2024 ; Tian et al. 2025 ). Recently, studies have shown that in the context of Massive Open Online Course (MOOCs) information overload of online learners experienced more anxiety and negative emotions maladapting their learning willingness and study adherence (Andrei Sitar-Tăut et al. 2024; Lv et al. 2024 ; Rahmani et al. 2024 ; Yusof et al. 2023 ). Overloaded and overwhelmed, a student may feel compelled to step away from the course to recover. In fact, the high quality online course often features with dense reading material and high demanding interactions often witness higher students dropout rates (Andrei Sitar-Tăut et al. 2024; Yusof et al. 2023 ). This echoes with students burn out or stressed in online coursework reports (Shi et al. 2020 ). Additionally, according to the digital disengagement literature, social and motivational factors are related with online learning. Comparing to physical teaching classroom, the students of online learning are tend to experience more isolated or unsupported feeling(Baltà-Salvador et al. 2021 ; Eberle and Hobrecht 2021; Li and Yang 2025). Loneliness is not only a byproduct of online learning, but also a emotional situation to trigger learning burnout (Li and Yang 2025). Learning burnout as a negative emotion features exhaustion, cynicism, and could lead to student academic engagement (Peng et al. 2023 ; Yusof et al. 2023 ). Li and Yang (2025) claimed that loneliness could act as a maladaptive factor of burnout among other factors (Li and Yang 2025). Concurrently, more and more online learner report the feeling of boredom and interest decreasing caused by insufficient peer interaction or feedback from instructors (Jimenez, Arguedas, and Borderia 2018). By the disengagement and interest declining are accumulated or added over time, it eventually becomes disengagement with behaviors of intermittent dropout away from or avoiding online class platform (Andrei Sitar-Tăut et al. 2024; Bekker et al. 2023; Yusof et al. 2023 ). Based on the literature, it can be founded that a combination of emotional stressors (anxiety, frustration, boredom) and contextual factors (isolation, overload, lack of support) are likely to integrated together leading to students learning intermittent dropout online (Yusof et al. 2023 ). Although there are many studies conducted on the issue of intermittent dropout, the work on systematically study dropout in technology-medicated context is limited (Rahmani et al. 2024 ). Previous studies typically investigated the individual predictors of online attrition (for instance, cognitive overload or loneliness or anxiety in isolation) by approaches of regression on correlated factors (Peng et al. 2023 ; Rahmani et al. 2024 ; Yusof et al. 2023 ). However, it is hard to justify the mechanism of student decision making from perspective of single factor, more recently there is a recognition that learners dropout is a combination of multiple converging factors and different students may leave for different combinations of reasons. Especially, in the environment of temporary disengagement and intermittent dropout, it is vital to explore how different factors jointly produce a tipping point where a student pauses or leaves a course. To answer this requirement, our study adopts a mixed method, configurational approach. We shall firstly investigate the student’s negative emotion mechanisms and their perceived triggers during episodes of disengagement. Then, using fuzzy-set Qualitative Comparative Analysis (fsQCA), we identify distinct configurations of factors that together lead to intermittent dropout. FsQCA is particularly suited to this investigation because it allows us to capture complex causality and equifinality (Ragin 2009 ) – the possibility that there are multiple, equally valid pathways to the outcome (dropout). We let patterns emerge from the data in a comparative case analysis; by doing so, this research strengthens the theoretical foundation of online student persistence by integrating emotional regulation perspectives with information overload and engagement frameworks. Practically, it yields insights into which combinations of issues (e.g. high workload and low support, or high anxiety and poor coping) are most dangerous, thereby informing more targeted interventions. In summary, this paper aims to unravel how negative emotions contribute to technology-mediated intermittent dropout through certain configurations of personal, emotional, and contextual factors. By following a clear mixed-methods design – qualitative exploration followed by fsQCA – we provide a well-rounded, evidence-based understanding of the dropout mechanisms. The article is structured as follows: first, we review relevant literature and theoretical frameworks on negative emotions, emotion regulation, overload, and disengagement in online learning. Next, we describe our methodology, including data collection, qualitative coding, calibration of variables, and fsQCA procedures. We then present the results, highlighting the key causal pathways identified. In the discussion, we interpret these findings in light of theory and prior research, discuss implications for educators and course designers, and acknowledge limitations and directions for future work. 2 Literature Review 2.1 Negative Emotions and Online Student Dropout Emotions are critical to the learning experience, and negative emotions in particular can undermine motivation and persistence (Li and Yang 2025; Miranda, Tulabut, and Cruz 2024; Wu and Liu 2024). In traditional classrooms, emotions like anxiety or frustration are transiently observable and can be addressed through in-person support (Álvarez et al. 2024 , 2024 ; He et al. 2024 ). In online settings, however, such emotions often accumulate unnoticed until a breaking point. Recent studies have solidified the link between negative academic emotions and dropout intentions (Cobo-Rendón et al. 2023 ). For instance, Cobo-Rendón et al. ( 2023 ) found that while positive emotions (enjoyment, hope) correlated with better adaptation and continued study, negative emotions were strong predictors of students’ intentions to drop out of university. Notably, anxiety – an emotion frequently reported in online courses – has been specifically implicated in higher dropout rates (Peng et al. 2023 ). In a German university study, students with greater academic anxiety were significantly more likely to discontinue their studies (Cobo-Rendón et al. 2023 ). These findings align with the broader understanding that emotions can drive academic behavior: students overwhelmed by negative feelings often disengage as a form of escape or self-preservation (Beard 2025 ; Xue 2023 ). Common negative emotions in online learning include frustration, boredom, anxiety, and hopelessness (Cloude et al. 2021 ; MacIntyre, Gregersen, and Mercer 2020 ; Solhi M. 2024 ). Frustration may arise from technical issues (e.g. platform glitches, poor internet connectivity) or confusing course materials (Cobo-Rendón et al. 2023 ). Boredom often stems from a lack of interactivity or monotonous content delivery (Li and Yang 2025). Anxiety is frequently tied to high-stakes assessments, fear of falling behind, or simply the self-directed nature of e-learning which can cause students to worry if they are learning effectively (Liu and Liu 2025 ). Prolonged exposure to these aversive emotions can lead to what Li and Yang (2025) term learning burnout, defined as a state where students feel exhausted, detached, and ineffective in their learning (Li and Yang 2025). Burnout is essentially an amalgamation of chronic negative feelings, and it has been shown to seriously reduce student engagement in online instruction (Bekker et al. 2023). A burnt-out student often exhibits apathy (lack of interest in course content) and diminished participation, behaviors that are on the spectrum of disengagement and precursor to dropout (Peng et al. 2023 ). It is important to note that negative emotions do not operate in isolation; they are often symptomatic of underlying problems in the learning environment (Eberle and Hobrecht 2021; Eres et al. 2021 ; Koundal et al. 2024 ). For example, a student might feel overwhelmed and anxious because the course workload is too high or instructions are unclear (He et al. 2024 ; Tian et al. 2025 ). Another might feel frustrated and helpless due to lack of timely feedback on assignments (Bahtilla 2024 ). A sense of boredom or pointlessness might actually reflect insufficient interactivity or real-world relevance in the course (Ahmed et al. 2024 ). Therefore, while we identify negative emotions as critical components in dropout cases, we must also examine their antecedents and context. By doing so, we can move beyond saying "anxious students drop out" to understand why the student became anxious to begin with (e.g. due to information overload or low self-efficacy) and how those factors combine to push the student toward disengaging. 2.2 Emotion Regulation in Online Learning Not all students who experience negative emotions will drop out; the difference often lies in how they regulate and cope with those emotions (Behr et al. 2020 ; Yusof et al. 2023 ). Emotion regulation theory, as developed by Gross and others, provides a framework for understanding this process (Li and Yang 2025; Liu and Liu 2025 ). According to Gross’s process model (Gross 1999 ), individuals deploy strategies either before an emotion fully unfolds (antecedent-focused strategies) or after the emotion has already been generated (response-focused strategies). In an academic context, an example of an antecedent-focused strategy is cognitive reappraisal – deliberately reframing a challenging situation in a more positive or neutral light (e.g., viewing a low quiz score as a learning opportunity rather than a catastrophe) (Zhao et al. 2021 ). A response-focused strategy example is expressive suppression – concealing or inhibiting the outward signs of frustration or panic, without addressing the root cause of the emotion (Heng et al. 2024 ). Each strategy impacts learning differently. Empirical evidence shows that adaptive emotion regulation is associated with better learning outcomes in online education. Zhao et al. ( 2021 ) demonstrated that students who habitually used cognitive reappraisal reported higher perceived control over their remote learning situation and ultimately learned more, whereas those who relied on suppression experienced heightened anxiety and learned less (Zhao et al. 2021 ). Crucially, anxiety was found to be inversely related to learning in that study. This implies that failing to manage anxiety can directly impair performance and satisfaction, potentially prompting students to withdraw (Bekker et al. 2023; Rahmani et al. 2024 ). In our context, this suggests that a student with good coping skills might endure a stressful online course without dropping out, whereas a student who cannot manage the stress might succumb to intermittent dropout. Maladaptive regulation can also accumulate feelings of isolation and distress (Heng et al. 2024 ). As noted earlier, suppressing emotions (a common reaction for students hesitant to seek help) can lead to a buildup of negative affect (Eden et al. 2020 ). Research during the pandemic observed that students who suppressed their worries were more likely to feel lonely and emotionally overwhelmed, compared to those who expressed and addressed their feelings (Li and Yang 2025; Preston et al. 2022 ). Loneliness and anxiety reinforced each other in a vicious cycle. Ultimately, such students may disengage as a coping mechanism, essentially removing themselves from the source of stress (the course) because they see no other way to obtain relief (Xue 2023 ). Literature on self-regulated learning also intersects here. Emotion regulation is one facet of the broader ability of students to manage their learning process (Eres et al. 2021 ; Gross 1999 ). High-performing online learners often exhibit self-regulation skills: they set goals, manage time effectively, seek help proactively, and keep their motivation up (Chang and Yang 2023). These behaviors help deepen negative emotions (for instance, good time management can prevent last-minute panic). Conversely, students with poor self-regulation may quickly feel overwhelmed, which can spiral into frustration or despair (Abdelhalim 2024 ). In our study, we consider emotion regulation capacity as a potential condition influencing dropout. We posit that how a student handles negative feelings (by reappraising challenges or by internalizing distress) could be a differentiator in whether they drop out intermittently when faced with difficulties. 2.3 Information Overload and Digital Disengagement The design and delivery of online courses can carelessly contribute to negative emotions through information overload (Tzafilkou, Perifanou, and Economides 2021 ). Online learning requires students to process large amounts of digital information daily – readings, videos, discussion forums, announcements, and more (35). When this volume of content exceeds a student's cognitive processing capacity or available study time, the result is often stress, confusion, and diminished learning (Li and Yang 2025). Chen et al. (2024) investigated high-quality MOOCs and found an intriguing paradox: courses with richer content and more interactive features (generally markers of quality) saw higher dropout rates, presumably because excessive content and interactions overwhelmed learners (Zhang and Chen 2021). They note that attention is a limited resource and scattering it across too many materials or activities can hinder learning outcomes. In practical terms, a student might start skipping videos or ignoring discussion boards to cope with overload, which then snowballs into disengagement (Kelly 2023 ). Stress and perceived difficulty from overload can manifest as anxiety and frustration, emotions strongly linked to dropout as discussed earlier. Digital disengagement is a broader concept capturing the ways learners withdraw from online platforms, whether temporarily or permanently (Bergdahl 2022 ). Causes of disengagement are multifaceted. A systematic review by Rahmani et al. ( 2024 ) highlights factors such as screen fatigue, where prolonged time in front of a computer leads to exhaustion and loss of focus, and feelings of isolation, where students miss the social dimension of learning (Rahmani et al. 2024 ). The rapid transition to online learning in the pandemic underscored these issues: many students reported boredom with online lectures, lack of motivation, and difficulty concentrating amid home distractions (Derakhshan et al. 2021 ). Additionally, lack of prompt support (unable to get immediate clarification from an instructor) often left students feeling lost (Beard 2025 ). These conditions contribute to what some authors call "digital fatigue" or "e-learning fatigue," essentially a state of weariness and disillusionment with online study. When students disengage digitally, it can take forms such as: not logging into the Learning Management System (LMS) for days or weeks, “ghosting” group projects, or ceasing to submit assignments – all without formally un-enrolling from the course (Bergdahl 2022 ). This intermittent dropout pattern can be seen as a coping response; students step away hoping to return once stress is lower or circumstances improve. Unfortunately, re-engaging after a long break is difficult, and many never fully catch up, leading to eventual failure or withdrawal. Previous research indicates that once disengaged, students face a high barrier to re-entry, especially if the course has moved on without them (Xue 2023 ). For example, if 75% of students disengage early as noted in the German survey, only a fraction of those are likely to return and complete the course. Thus, preventing disengagement in the first place is critical. In summary, information overload and related design issues in online courses can evoke the very negative emotions that drive students away (Shi et al. 2020 ; Tian et al. 2025 ). Coupled with insufficient support or social interaction, these factors create a trigger for digital disengagement (Xue 2023 ). The literature underscores an important point: any successful intervention to reduce online dropout must address not just technical or academic preparation, but also the emotional and cognitive load placed on students. Reducing unnecessary complexity, balancing content quantity, fostering community, and providing timely help are all recommended practices to mitigate overload (Tian et al. 2025 ). Our study builds on this literature by examining how overload, isolation, and other factors combine with personal emotions in actual student dropout cases. 3 Conceptual Framework and Research Questions Bringing together the above strands, we conceptualize intermittent dropout in technology-mediated learning as an outcome emerging from multiple interacting factors. Figure 1 illustrates the proposed framework. In brief, environmental stressors (like heavy workload, confusing UI, or technical problems) can lead to information overload, which in turn provokes negative emotional responses (e.g., anxiety, frustration). Simultaneously, social/contextual deficiencies (like lack of instructor feedback or peer interaction) contribute to feelings of isolation or boredom, another set of negative emotions. The student's personal characteristics (such as emotion regulation ability, self-efficacy, time management skills) moderate how these experiences are internalized – a resilient student might cope, whereas a vulnerable student might accumulate stress. When certain combinations of these conditions occur (e.g., high overload, low support, poor coping), the student crosses a threshold and disengages temporarily (an intermittent dropout event). What remains unclear, and forms the gap we address, is the exact combination of factors that lead to such dropout events. Traditional quantitative methods might tell us that "overload increases odds of dropout by X%" or "anxiety correlates with dropout intentions," but they do not illuminate how factors cluster in individual cases (Rahmani et al. 2024 ; Yusof et al. 2023 ). By using fsQCA, we approach this as a configurational problem: identifying sets of conditions that are sufficient to produce the outcome. This approach aligns with recent calls to use configurational comparative methods in education research to capture complex causality. While a few studies have applied QCA to topics like e-learning extension or student satisfaction, to our knowledge none have focused on emotional mechanisms in dropout behavior. Given the gap, our study raise questions: What configurations of negative emotions, personal factors, and contextual conditions lead to technology-mediated intermittent dropout? What is the mechanism of student negative emotions in technology-mediated intermittent dropout? We expect, based on the review above, that conditions such as high perceived workload, feelings of isolation, strong negative emotions (anxiety, frustration), and poor coping ability will feature prominently in the dropout pathways. Conversely, we anticipate that absence of these issues or presence of protective factors (e.g., high support, effective emotion regulation) will distinguish those who persist. These expectations inform our analysis; the fsQCA will reveal which combinations actually mattered for our participants. By addressing this research objective, our study contributes to theory by integrating emotional and cognitive perspectives on dropout, and to practice by identifying leverage points (which factor combinations to target) to reduce online student attrition. The next section details how we designed the study and carried out the mixed-methods investigation to achieve these aims. 4 Methodology 4.1 Research Design We employed a mixed-methods sequential exploratory design consisting of a qualitative phase followed by a quantitative fsQCA phase. The rationale for this design was to first gain an in-depth understanding of students' emotional experiences and the context of their intermittent dropout (through qualitative inquiry), and then to systematically analyze patterns across cases using fsQCA. This approach allowed us to ground the quantitative analysis in real-world insights and to ensure that the conditions fed into the fsQCA model were relevant and evidence-based. The design can be summarized in two broad stages: Qualitative Exploration: We conducted semi-structured interviews with students who had experienced intermittent dropout in an online learning context. The goal was to uncover which factors (emotional, academic, environmental, etc.) they perceived as contributing to their disengagement episodes, and how those factors interrelated. We also probed how students responded to challenges (coping strategies) to assess their emotion regulation in context. Configurational Analysis (fsQCA): Based on the qualitative results, we identified key conditions and calibrated these as fuzzy-set variables. Each student's case (from the interviews, and supplemented by additional data where available) was then represented as a case in a comparative analysis. Using fsQCA software, we examined which combinations of conditions were present in cases where dropout occurred, thereby identifying sufficient causal recipes for intermittent dropout. This design aligns with fsQCA's strengths in handling complex causality and small-to-medium sample sizes, while the qualitative component enhances the study's credibility and contextual richness. By clearly explaining each step – from qualitative coding to variable calibration to solution interpretation – we heed the reviewer's call for a transparent research process. Ethical approval was obtained from the Tongling Polytechnic Academic Committee prior to data collection. All participants gave informed consent, and their identities were kept confidential (pseudonyms or ID codes were used in lieu of real names in transcripts and analysis). Participation was voluntary, and students were assured that declining or withdrawing would have no effect on their academic standing. These measures ensured an ethical and open environment for participants to share their experiences of dropout. 4.2 Qualitative Phase: Interviews and Coding Participants The qualitative phase involved N = 20 participants recruited from a large online degree program at a university (or a MOOC platform, etc.). We used purposive sampling to select students who had shown evidence of intermittent dropout, defined operationally as having temporarily stopped engaging with an online course for a significant period (e.g., missing several weeks of course participation or assignments) before either returning or eventually withdrawing. The sample (detailed in Table 1 ) included a mix of genders, and majors to capture diverse perspectives. The mean age was roughly 20 (with a range from late teens to mid-career adults, reflecting typical online learner demographics). We stopped recruitment at 20 interviews as we reached thematic saturation (no fundamentally new themes were emerging in later interviews). Table 1 Participant Information Statistics Table Category Description Percentage/Number Gender Female 49.1% (n = 146) Male 50.9% (n = 150) Age Distribution 18–24 years old 100% (n = 296) Educational Background College students 100% (n = 296) Frequency of Educational Technology Use Daily use 70 participants (23.64%) Several times a week 186 participants (62.41%) Once a week or less 40 participants (13.51%) Experience with IDB Intermittent Discontinuance Behavior 214 participants (72.29%) Anxiety 83.6% Dissatisfaction 76.9% Depression 68.2% Worries 71.5% Data Collection We developed a semi-structured interview guide with open-ended questions covering areas such as: students’ overall online learning experience, specific instances of disengagement (e.g., "Can you describe a time when you stopped participating in an online course? What was happening at that time?"), emotions felt during that period (prompting for feelings like stress, frustration, boredom, etc.), perceived causes or triggers of those emotions, support or lack thereof from instructors/peers, and any strategies they used to cope or re-engage. Follow-up questions probed deeper into key statements (for example, if a student said "I was completely overwhelmed," we asked what contributed to that feeling). Interviews were conducted via video conferencing and lasted approximately 25–40 minutes each. All interviews were audio-recorded (with permission) and transcribed verbatim for analysis. Qualitative Analysis We used a thematic analysis approach (following Braun & Clarke's guidelines) to code and interpret the interview data. Initially, two researchers independently read through a subset of transcripts to familiarize themselves with the content and note preliminary ideas. Next, they collaboratively developed a coding scheme that balanced deductive codes (derived from our conceptual framework, e.g., "technical difficulties", "workload", "anxiety", "seeking help") and inductive codes (new themes emerging from the data, e.g., "family responsibilities" as a reason to drop out, if it arose). Each transcript was coded line-by-line in qualitative analysis NVivo software. The two researchers double coded about 25% of the transcripts to ensure reliability, achieving an inter-coder agreement of ~ 0.85 (Cohen’s kappa) on key thematic categories after resolving minor discrepancies through discussion. Through this coding process, we identified recurrent factors associated with the students’ dropout episodes. These factors (which later informed our fsQCA conditions) included, for example: Overwhelming Workload/Content: Many students described the course content as too dense or fast-paced, using phrases like "information overload" or "swamped with work." This often led to feelings of anxiety and helplessness. Technical or Logistical Issues: Some cited persistent technical breakdown, unreliable internet, or difficulty using the learning platform, which caused frustration and disrupted their learning routine. Lack of Instructor Support: A common theme was that when students struggled, they did not get timely help or feedback. Emails or forum questions went unanswered, leaving them feeling isolated ( "I was on my own" ). Low Peer Interaction: Especially during pandemic remote learning, students missed having classmates to study with or simply to commiserate. The absence of a learning community contributed to boredom and a sense that “no one would notice if I disappeared.” Personal Life Stressors: External factors like work commitments, family responsibilities, or health issues sometimes precipitated disengagement. These were often accompanied by stress and guilt, and if the course was not flexible, the student would drop out to cope. Negative Emotional States: Across nearly all interviews, students explicitly mentioned emotions – frustration (“I got so frustrated that I gave up for a while”), anxiety (“I was constantly anxious about falling behind”), boredom (“The class was just so dull, I lost motivation”), and shame or low self-efficacy (“I felt I wasn’t smart enough to do the assignments”). These emotions were described both as outcomes of the above factors and as immediate triggers for stepping away. Coping and Regulation: We observed differences in how students handled their emotions. Some tried positive coping (contacting the instructor, taking a short break and coming back), whereas others resorted to avoidance (completely logging off, ignoring the problem). Notably, very few had formal strategies to regulate emotions; most actions were ad-hoc. Students who eventually returned to the course sometimes cited an external push (deadline pressure, a reminder from a friend, etc.) rather than intrinsic regulation. Through iterative discussion and concept mapping, we synthesized these findings into a set of candidate conditions to be used in the fsQCA. We ensured each condition was clearly defined and grounded in multiple interviewees’ accounts (not a one-off mention). For example, “Overload” was defined as the student feeling the course demands exceeded their capacity, evidenced by quotes about excessive content or time pressure; “Lack of support” was defined by instances where students could not get help when needed; “High negative emotion” was defined by intense feelings like severe anxiety or frustration that the student linked to their dropout decision. Importantly, we also noted cases of non-dropout within our sample (some students described difficult situations which they managed to overcome and did not drop out). These provided contrastive insights – e.g., a student who struggled but persisted often had at least one mitigating factor (like strong self-motivation or external support). The outcome of the qualitative phase was thus a nuanced understanding of the mechanisms connecting various factors to negative emotions and dropout. We compiled a summary for each interviewee (case) detailing whether each of the identified factors was present or salient in their story. This case-wise data became the basis for calibrating fuzzy-set membership scores in the next phase. 4.3 Calibration of Variables for fsQCA Using the qualitative insights, we proceeded to operationalize the conditions and outcome for the fsQCA. FsQCA requires each case (here, each student) to have a score between 0 and 1 for each condition and the outcome, representing degree of membership in the set (e.g., the set of students experiencing overload, or the set of students who dropped out). We defined the following key outcome and conditions detailed in Table 2 : Table 2 Calibration Table Variables Fully Membership (0.0) Partial Membership (0.5) Fully Non-Membership (1.0) Intermittent Dropout Clear episode of disengagement Reduced activity, borderline case Non-dropout Overload Strongly felt overload Moderate overload at times Never mentioned workload Lack of support Felt adequately support Minimal or delayed support Experienced major support gaps Negative Emotions Not report Moderate negative emotions present Strong negative emotions described Poor Emotional regulation Proactive and adaptive coping Partial Not cope or maladaptive Isolation Felt well-connected Some interaction insufficient Felt very isolated Outcome (Intermittent Dropout) : This was the target set we are trying to explain. We coded a case as 1 (full membership) in dropout if the student had a clear episode of disengagement (e.g., stopped participating for a significant period or officially withdrew) during the course in question. If a student did not disengage at all (continuous participation), they would be 0 (full non-membership). Partial membership (0.5) could be assigned in borderline cases – for instance, if a student reduced activity for a while but did not completely stop (though we tried to keep outcome binary for interpretability). In our data, since we specifically sampled dropout cases, most cases were 1; however, we included a few non-dropout cases (students who nearly dropped but managed to continue) as counterexamples to enrich the analysis. Condition 1: Overload (OL) – Reflecting information overload or overwhelming workload . Using interview evidence and any available performance data (like hours spent or number of assignments overlapping), we calibrated this as: 1 if the student strongly felt overloaded (explicit statements of being overwhelmed, consistently unable to keep up with content), 0.5 if moderate (some overload at times, but manageable or mixed experience), and 0 if overload was not a factor (student never mentioned workload as an issue or explicitly said workload was fine). Condition 2: Lack of Support (LS) – Capturing insufficient instructor or institutional support . We gave 1 if the student experienced major support gaps (e.g., no response when seeking help, felt completely unsupported), 0.5 if support was minimal or delayed (e.g., help came but too late or only after repeated requests), and 0 if the student felt adequately supported (instructor was responsive, resources were sufficient). Condition 3: Negative Emotions (NE) – Indicating the presence of intense negative emotional experience during the course. We focused on emotions like anxiety, frustration, and demotivation. 1 meant the student described strong negative emotions that significantly affected them (e.g., panic attacks, extreme frustration leading to anger or tears), 0.5 for moderate negative emotions (stress was present but perhaps not incapacitating, or fluctuated), and 0 if the student did not report notable negative feelings (this was rare in our dropout cases, more common in non-dropouts). Condition 4: Poor Emotion Regulation (PR) – Representing ineffective coping or regulation strategies . This was a more inferential measure based on interview responses to how they dealt with challenges. 1 was assigned if the student basically did not cope or used maladaptive strategies (withdrew completely, ignored issues, self-blame without action), 0.5 if they made some effort but it was partial (e.g., took a break but maybe too late, or tried one strategy that didn’t fully work), and 0 if they demonstrated proactive and adaptive coping (sought help, adjusted study strategies, used positive reframing). We cross-validated this with their outcome – often, dropouts had high PR (poor regulation) whereas those who navigated difficulties had low PR (good regulation). Condition 5: Isolation (IS) – Reflecting low social presence or peer engagement . 1 if the student felt very isolated (no peer interaction, explicitly complained of loneliness or lack of collaboration), 0.5 if there was some interaction but insufficient (maybe they had group work but it was perfunctory, or they felt “disconnected” even though others were technically present), and 0 if the student felt well-connected (had active communication with peers or instructors). This condition is tied to disengagement from the community aspect of learning. (Additional conditions could include personal factors like self-efficacy or external pressures, if our qualitative data strongly indicated them. For brevity, we list five above, which were among the most recurrent.) Each case (student) thus could be represented as a vector of these scores. We employed the direct method of calibration, where we set three qualitative anchors for each fuzzy set: one for full membership (criterion for being fully in the set), one for the crossover point (0.5, the point of maximum ambiguity where the case is neither in nor out), and one for full non-membership. These were based on our code definitions and, where possible, numeric indicators. For example, for Overload , if a student explicitly said "I had no trouble with the workload," that was evidence for a score near 0; if they said "I was drowning in readings and assignments," that was evidence for a score of 1. Intermediate statements (or mixed signals) would lead to a 0.5. By anchoring the calibration to qualitative evidence (often using participant quotes as benchmarks), we ensured our fuzzy-set scores stayed true to the data (a process akin to “anchored calibration” in QCA methodology). To enhance reliability, a second researcher reviewed the calibration of each case; disagreements (e.g., whether a certain description warranted a 0.5 or 1) were resolved through discussion, referring back to transcripts. We used fsQCA software (Version 3.0) to input the calibrated data for analysis. Before constructing the truth table, we checked for necessary conditions. A necessity analysis assesses if any single condition is present in all or nearly all instances of the outcome (i.e., does any factor appear to be a prerequisite for dropout?). We found no condition with a consistency above the typical threshold of 0.90 for necessity – in other words, no single factor was absolutely required for an intermittent dropout to occur. This reinforces the idea that different combinations can lead to dropout (equifinality), and it justified our focus on sufficiency analysis via the truth table. 4.4 fsQCA Procedure and Solution Derivation With calibrated data ready, we proceeded to the core fsQCA analysis: Truth Table Construction The software generated a truth table, which lists all possible combinations of the conditions (2^5 = 32 possible rows, for five conditions) and indicates which combinations are present in our observed cases and whether they lead to the outcome. Given our sample size (~ 20 cases), many logically possible combinations had no cases (so-called “logical remainders”). We set a frequency threshold of 1 case per combination to consider a configuration empirically relevant (since our sample is small, even a configuration with 1 case is of interest, but in larger samples one might set this higher). We set a consistency threshold of ~ 0.80 for a combination to be considered sufficient for the outcome. Consistency here measures how uniformly cases with that combination exhibit the outcome – values close to 1 indicate that whenever that combination is present, the outcome is almost always present as well. We examined the truth table rows exceeding 0.8 consistency and flagged them for inclusion in the logical minimization. Logical Minimization: Using QCA’s minimization algorithm (Quine–McCluskey), we derived simplified solution terms that explain the outcome. We opted to derive the intermediate solution, which balances between the complex solution (no remainders used, very conservative) and the parsimonious solution (maximal use of remainders, risk of over-simplification). To produce an intermediate solution, one must specify directional expectations for how each condition relates to the outcome (i.e., whether the presence or absence of each condition is theorized to contribute to dropout). Based on literature and our qualitative insights, we assumed that the presence of risk factors (overload, lack of support, strong negative emotions, poor regulation, isolation) would favor the occurrence of dropout, whereas their absence or opposites would favor persistence. These assumptions guided the inclusion of logical remainders that did not contradict our theoretical understanding. Importantly, we kept these expectations moderate. The software then provided the intermediate solution, consisting of one or more configurations (causal recipes) that are sufficient for the outcome. Solution Evaluation: We recorded the consistency and coverage of each configuration and of the overall solution. Consistency of each solution term indicates how reliably that combination leads to dropout (again we sought values well above 0.8). Raw coverage tells us the proportion of outcome cases explained by that term, and unique coverage indicates the proportion explained only by that term (i.e., cases not covered by other terms). We also looked at the solution consistency and coverage as a whole. A high solution consistency (e.g., > 0.8) means the set of identified configurations collectively are good predictors of dropout cases, and a decent solution coverage (say 0.5–0.7) means they explain a substantial fraction of all dropout cases, recognizing that some cases might remain unaccounted for if they had very idiosyncratic causes. Robustness and Sensitivity Checks: To ensure our findings were not an artifact of specific parameter choices, we conducted several checks: We raised the consistency threshold to 0.85 to see if the solutions changed; the core configurations remained the same, giving confidence in their robustness. We tried calibrating the outcome as a fuzzy set (for those who partially disengaged) instead of a crisp set; this did not substantially alter which conditions appeared important. We ran an analysis of the negation of the outcome (i.e., examining configurations for students who persisted without dropout) to see if it yielded simply the mirror image of dropout configs or something different. This revealed, interestingly, that persistence often required the absence of multiple risk factors simultaneously (no overload and good support and so on), reinforcing that dropout can be triggered by the presence of any one pathway of problems, whereas staying engaged may demand that several things go right at once. We also tested removing one condition at a time to check if any single condition was overly influential (e.g., if removing "Negative Emotions" drastically drops solution consistency, it means most paths relied on that condition). We found that while some conditions (like NE) appeared in multiple solutions, the overall phenomenon could still be explained in their absence via other routes, confirming that multiple independent pathways exist. Throughout the analysis, we prioritized a transparent and data-driven approach . All analytic decisions (calibration thresholds, consistency cut-off, inclusion of remainders) were documented and can be traced back to either empirical evidence or theoretical rationale. By avoiding arbitrary or biased choices, we aimed to let the data “speak for itself,” fulfilling the review requirement to let conclusions emerge from the fsQCA rather than imposing preconceived hypotheses. In the next section, we present the results of the fsQCA, including the specific configurations identified as leading to intermittent dropout, illustrated with examples from the qualitative cases to give them concrete meaning. 5 Results 5.1 Qualitative Insights: Factors Precipitating Dropout Before delving into the fsQCA configurations, it is helpful to briefly summarize the qualitative insights that set the stage. As noted, the interviews revealed a constellation of factors that frequently co-occurred with students’ decisions to temporarily drop out. To illustrate, we highlight a few anonymized cases: Case A (Overload & Anxiety): A first-year student described how a combination of five challenging courses and continuous weekly assignments led to a crushing workload. She recounted, “Every day there was something due. I couldn’t keep track.” This overload triggered intense anxiety and sleepless nights. Without an outlet or guidance on managing the load, she stopped logging into two of her courses for a month (an intermittent dropout) to focus on her mental health. Case B (Isolation & Boredom): Another student, enrolled in a fully online program during the pandemic, felt extreme isolation. He said, “I never really got to know my classmates. Discussion boards felt like talking to a void.” Lacking social interaction, he lost interest and described the course as boring. Midway, he disengaged for several weeks, saying “I figured it didn’t matter to anyone if I showed up or not.” Case C (Tech Issues & Frustration): A working adult student faced recurrent technical problems – the learning platform would crash during quizzes, and video lectures were often inaccessible on her network. These issues led to mounting frustration: “I was spending more time troubleshooting than learning.” After failed attempts to get IT support, she temporarily dropped out in exasperation. Case D (Personal Stress & Lack of Support): A part-time student balancing family responsibilities had a medical emergency at home. She informed the instructor she would be offline for a week, but got no extension or meaningful help catching up. Feeling overwhelmed by personal stress and unsupported academically, she disengaged. She later said, “If the professor had just reached out or given me some leeway, I might have continued.” These examples show how multiple factors converged in each dropout episode. Importantly, not every case had all factors present; rather, different subsets of factors drove different students to disengage. This observation is what the fsQCA results formalize. 5.2 Configurations Leading to Intermittent Dropout (fsQCA Findings) The fsQCA yielded several configurations (causal pathways) that were sufficient to produce the outcome of intermittent dropout. Table 1 below summarizes the simplified solution terms obtained, with each configuration indicating a combination of conditions. (Note: In a text format here, we describe them in prose, but one could imagine a table with conditions marked present (●), absent (⊗), or irrelevant for each solution, alongside consistency and coverage metrics.) Table 3 Configuration path leading to Intermittent Dropout Variables Solution Cases Path 1 Path 2 Path 3 Path 4 Overload ● ⊗ ⊗ ⊗ Lack of Support ● ⊗ ⊗ ⊗ Negative Emotions ● ● ● ⊗ Isolation ⊗ ● ⊗ ⊗ Poor Emotion Regulation ⊗ ⊗ ● ⊗ Raw Coverage 0.45 0.30 0.25 Unique Coverage 0.013 0.011 0.06 Consistency 0.92 0.88 0.85 Solution Coverage 0.849 Solution Consistency 0.811 Solution 1: Overload AND Lack of Support AND Negative Emotions This configuration can be expressed logically as: OL * LS * NE (meaning overload, combined with lack of support, combined with presence of strong negative emotions). Consistency = 0.92, Raw coverage = 0.45. In plain terms, Solution 1 indicates that when students experience a heavy workload or information overload (OL = 1) together with insufficient support from instructors (LS = 1), and as a result they suffer intense negative emotions like anxiety or frustration (NE = 1), they are very likely to drop out intermittently. Several cases aligned with this pattern. For example, in Case A above, we see exactly this: enormous content demands plus no felt support, yielding severe anxiety. Most students under this triple strain reached a breaking point. Notably, this path reflects a classic overload-burnout scenario: the environment overloads the student, the student doesn't get help, and their emotional state collapses, forcing disengagement. Solution 2: Isolation AND Boredom (or Lack of Interest) This configuration, simplified, corresponds to IS * (lack of engagement). In terms of our conditions, that could be captured by IS = 1 (isolation present) and perhaps also NE = 1 if boredom is considered a negative emotion. In our formal model, boredom was one aspect of NE, but here it's specifically tied to isolation. Consistency = 0.88, Raw coverage = 0.30. Solution 2 suggests that social isolation, coupled with a resulting loss of interest or boredom, is another pathway to dropout. This was evident in Case B: feeling no social connection (IS) led to disengagement out of apathy. Interestingly, in some of these cases the workload was not necessarily high – it was the lack of meaningful interaction that made the course feel not worth pursuing. This configuration resonates with the concept that emotional disengagement (boredom, apathy) can be just as detrimental as cognitive overload. It also highlights the importance of social presence: in our data, no student who felt well-integrated socially dropped out purely from boredom, indicating isolation was a key ingredient. Solution 3: Poor Emotion Regulation AND High Negative Emotion (even if workload manageable) This path can be denoted as PR * NE, with absence of OL (i.e., ~OL) in some instances. Consistency = 0.85, Coverage = 0.25. It points out cases where a student’s inability to cope (PR = 1) with even moderate challenges led to runaway negative emotions (NE = 1) and eventually dropout. In some instances here, the objective course demands were not extreme (so overload was absent or low), but the student’s personal coping threshold was low. For example, a student might have had perfectionist tendencies or high anxiety trait; a small setback (like one poor grade) spiraled into major self-doubt and panic because they didn’t utilize healthy coping strategies (no reappraisal, only rumination). These students often withdrew even though external conditions alone might not predict it. Solution 3 underscores the role of individual differences in emotion regulation – even a well-designed course can lose students if they lack resilience or coping mechanisms. Conversely, it implies that building students’ self-regulatory skills could prevent dropout in cases where academic factors are surmountable. Solution 4: External Stressor AND Lack of Flexibility (a less frequent but present pathway): Although not among the top three configurations by coverage, we observed a pattern where an external life stressor (like health or family issues, not formally one of our main conditions but qualitatively noted) combined with an inflexible course structure (could be seen as a special case of lack of support or a high workload that can't be adjusted) led to dropout. This is effectively a conjunctural cause outside the core model but worth mentioning. Consistency was high for these few cases. It reminds us that sometimes the trigger to disengage comes from outside academia, but whether it results in dropout depends on the course’s accommodation (or lack thereof) for the student's crisis. Across these solutions, certain conditions stood out. High Negative Emotions (NE) was part of nearly every configuration – confirming that emotional distress is a common denominator in dropout pathways. However, as our analysis showed, NE alone was not sufficient; it was always accompanied by precipitating factors like overload or isolation. Lack of Support (LS) and Isolation (IS) each appeared in at least one major solution, indicating that either academic or social support deficits can facilitate dropout, albeit through different emotional routes (acute frustration vs. chronic apathy, respectively). Overload (OL) was critical in one of the highest coverage paths, aligning with the notion that excessive demands often directly push students away. Poor Regulation (PR), while harder to measure directly, emerged as an important internal condition explaining why some students quit even under not-so-extreme external conditions. The solutions also exhibit equifinality : Solution 1 and Solution 2, for example, describe two very different profiles of at-risk students (stressed-overwhelmed vs. isolated-bored). Both profiles can lead to dropout, implying that interventions need to be multifaceted. Another observation is the idea of conjunctural causation – it's the conjunction of factors that matters. Overload by itself was not enough to guarantee dropout if, say, support was present or the student coped well; isolation by itself might be endured if the student still found the content engaging or had strong personal motivation. It's when these factors co-occur without counterbalancing positives that dropout occurs. From a coverage standpoint, the combination of the identified solutions explained a majority of the dropout cases in our sample (solution coverage ~ 0.70, meaning 70% of dropout instances had at least one of these configurations present). A few cases remained outliers, which upon examination had very unique circumstances (for instance, one student dropped out mainly due to a sudden job offer – a case of opportunity rather than negative experience). Such cases remind us that no model captures 100% of human behavior, but they were exceptions. 5.3 Configurations for Persistence (Contrast Analysis) As a complementary analysis, we also briefly examined what configurations were associated with students persisting (not dropping out) despite challenges. While not the primary focus of our study, this provides a mirror image to interpret results. Persisting students tended to have at least one of the following protective patterns: strong support (~ LS = 0) or low workload (~ OL = 0) or effective coping (~ PR = 0) – and usually more than one of these in combination. In other words, to not drop out, it helped to have multiple positives: e.g., a manageable workload and good instructor support and perhaps only mild negative emotions. This asymmetry (dropout can be caused by any one severe pathway, whereas persistence often requires all-clear on many fronts) is a sobering insight. It highlights why dropout rates are often high – there are many ways to drop out (many causal recipes), but to stay consistently engaged, a student ideally needs a more complete support system and a bit of luck in avoiding major stressors. To summarize the results: our fsQCA identified multiple sufficient pathways to intermittent dropout, each involving a mix of negative emotions and contributing factors . These findings validate that no single factor is responsible for online dropout; instead, it is the confluence of academic, social, and emotional factors that creates the conditions for a student to disengage. In the next section, we discuss these findings in light of existing theories and research, and explore what they mean for educators and institutions aiming to improve student retention in technology-mediated learning. 6 Discussion and Implications 6.1 Discussion Our findings shed light on the mechanisms through which negative emotions drive students to intermittently drop out of online courses, and they reinforce the idea that dropout in technology-mediated environments is a configurational phenomenon . In this discussion, we interpret each major pathway identified, connect it to theoretical frameworks and prior research, and then delve into the broader implications for theory and practice. Pathway 1: Overload + No Support → Anxiety/Frustration → Dropout. This pathway aligns well with information overload theory and the concept of academic burnout . When students face an excessive volume of content or tasks without adequate guidance, they can experience cognitive overwhelm that quickly turns into intense anxiety and frustration. Our results echo the findings of Chen et al. (2024) who warned that too much content and interaction can backfire, causing stress and lowering learners’ willingness to continue . In our data, overload was not acting alone; it became truly perilous when combined with lack of support . This pairing fits with Digital Stress models: stressors (like overload) coupled with lack of resources (like support) yield strain (negative emotions) and eventual withdrawal (dropout). Theoretically, this underscores the importance of P-E fit (Person-Environment fit) in online learning – a mismatch between demands and support triggers negative affect and maladaptive outcomes. From an emotion regulation standpoint, even a student with decent coping skills can be overwhelmed if the objective demands far outstrip their capacity; no amount of deep breathing will help if you simply have 48 hours of work to do in 24 hours. Therefore, while we encourage teaching students coping techniques, the onus is also on course designers to manage workload and information flow to prevent such overload situations. Pathway 2: Isolation → Boredom/Disinterest → Dropout. This configuration highlights the social and emotional void that can occur in online education. It maps onto literature about emotional engagement and belonging : students who feel disconnected from peers and instructors often fail to develop investment in the course. Over time, this becomes boredom and apathy , which are emotions strongly linked to disengagement. Prior research by Li and Yang (2025) found that loneliness in online learning is a significant contributor to burnout; our findings deepen that by showing loneliness (isolation) can specifically breed boredom and loss of purpose, precipitating dropout. This resonates with the self-determination theory notion that relatedness is a basic human need – when not met, motivation suffers. In on-campus settings, even a dull lecture can be made bearable by the presence of friends or the college environment; online, a dull course with no community feels especially pointless. Importantly, this pathway did not necessarily involve high workload or intense anxiety – it was a slow erosion of engagement rather than an acute crisis. The implication for theory is that dropout can stem from absence of positive emotion or stimulation , not just presence of negative stress. Practically, it calls for integrating community-building and interaction in online courses. Even simple interventions like prompt instructor feedback, synchronous meet-ups, or group projects could mitigate feelings of isolation. Research shows that fostering social presence can significantly improve engagement and reduce attrition, as “social presence relieves burnout” by alleviating loneliness. Our results strongly support that: none of the well-connected students followed this boredom path. Pathway 3: Poor Emotion Regulation → Unchecked Negative Emotions → Dropout. This highlights the role of the student’s internal capacities . Even when external conditions were not extreme, some students fell into a dropout pattern because they couldn’t manage moderate challenges – small setbacks snowballed. This finding speaks directly to emotion regulation theory : those who lacked effective coping strategies (e.g., they engaged in rumination, avoidance, or denial) allowed their negative emotions to accumulate unchecked. The case of a student catastrophizing a single bad grade into “I’m not cut out for this” is a prime example. Zhao et al. ( 2021 ) demonstrated that students using maladaptive strategies like expressive suppression ended up with more anxiety and poorer outcomes, which dovetails with what we saw: suppressors (or generally poor regulators) in our sample often ended up disengaging because their anxiety or frustration became overwhelming. This pathway suggests that personal resilience factors are an important part of the dropout equation. It also partially answers why some students weather the same storm that sinks others. From a theoretical viewpoint, it invites a more nuanced model of online persistence that incorporates emotional self-regulation as a moderator between challenges and outcomes. Interestingly, our configurational approach suggests that improving emotion regulation alone might prevent certain dropouts, but not all – because some paths (like the overload one) might overpower even good regulators. Conversely, even a perfectly designed course can lose students who are unprepared to deal with any emotional discomfort. Thus, responsibility is dual : institutions should train students in coping and mindset (for example, workshops on time management, stress reduction, fostering a growth mindset to handle failure) and simultaneously design courses that are supportive and humanized to reduce undue stress. This dual approach tackles both sides of the emotion equation. Our results empirically back the argument made by some educators that “we need to teach emotional skills, not just content” – something especially pertinent in self-driven online learning. Pathway 4 and others: Additional nuances. We noted an external-stressor path (life event + inflexibility) which aligns with research noting that many online learners are adults juggling multiple roles; life often intervenes. While not a primary focus, it reminds us that structural flexibility (e.g., self-paced options, lenient deadlines in genuine emergencies) can catch those falling due to outside reasons. Another pattern in persistence we observed is that to avoid dropout, often multiple positive factors must coincide (the student must not be overloaded, and feel supported , and manage emotions well, etc.). This asymmetry is worth reflecting on: it implies that preventing dropout is harder than causing it , because one weak link in the chain (one risk factor) can trigger disengagement, whereas preventing it requires shoring up all links. This may explain why dropout rates remain high – it's easier for something to go wrong than for everything to go right. The implication is that institutions should implement redundant retention strategies : assume that just fixing one issue (say, providing tutors) may not be sufficient by itself; a combination of improvements (better course design, better support, student skills training, etc.) is needed to truly curb dropout rates. Comparison with Prior Studies : Our results both confirm and extend prior empirical findings on online dropout. The systematic review by Rahmani et al. ( 2024 ) identified numerous factors like anxiety, isolation, and workload issues as contributors to dropout. We corroborate those factors and show how they interlink . For instance, Rahmani et al. noted anxiety and concentration problems during the pandemic as major issues; our overload pathway is one concrete route to anxiety-driven dropout. Respondek et al. ( 2017 ) found anxiety predictive of dropout in traditional university; we again echo that but embed it in a causal combination. Our study also relates to research on academic emotions : Pekrun’s control-value theory suggests that emotions like boredom and anxiety arise from students’ appraisals of a situation (value and control). Our isolation-boredom path suggests students saw little value (no social or intrinsic value) and possibly low control (stuck in a boring environment they couldn’t change), leading to disengagement – aligning with that theory. Meanwhile, the overload-frustration path can be seen through a control-value lens as well: those students likely felt low control (over the immense workload) and high value might not salvage them because they were simply incapable of controlling outcomes, resulting in anxiety (a negative activating emotion) which led to giving up. 6.2 Implications The insights from this study can inform several concrete actions for educators, instructional designers, and academic policy makers: Optimize Workload and Content Delivery : Given the clear risk of information overload, online courses should be carefully moderated in terms of content density and frequency of assessments. Quality should be favored over quantity. Instructors might break content into manageable chunks, provide clear guidance on study time expectations, and use adaptive release (staging content rather than dumping everything at once). Periodic check-ins can gauge if students feel overwhelmed. As Chen et al. (2024) note, more content is not always better – our data show it can be worse. Thus, content richness should be paired with equally rich support to avoid overload. Enhance Instructor Presence and Support : The lack of support factor implies that responsiveness is key. Simple practices like replying to questions within 24–48 hours, holding virtual office hours, and proactively contacting inactive students can make a difference. Some institutions employ early alert systems that flag when a student hasn’t logged in or submitted work; instructors (or support staff) can then reach out in a supportive, non-punitive way (“We noticed you haven’t been active, is everything okay? How can we help?”). Establishing this net can catch students on the brink of dropout. Additionally, providing timely and constructive feedback on assignments helps reduce anxiety and uncertainty about progress. Build Social Interaction and Community : To combat isolation and boredom, course designers should integrate discussion forums, group projects, peer feedback, or live webinars where students can see and hear each other. Even if a course is asynchronous, including an initial introduction forum, collaborative activities, or study buddy systems can foster peer connections. In our findings, those who had somebody – be it a friend, a mentor, or an engaged instructor – were far less likely to disappear silently. Institutions might also encourage formation of student-led study groups or use technology (like social learning platforms) to create a sense of campus-like community online. These efforts are supported by evidence that social engagement can dramatically increase completion likelihood (one stat suggests students engaged in communities are 5x more likely to complete the course – although context-specific, it illustrates the magnitude). Develop Students’ Emotional and Self-Regulation Skills : Many students enter online learning without prior experience in that mode, and they might lack the self-regulation habits needed. Orientation modules that teach time management, goal setting, and emotional self-awareness could be made standard. Furthermore, integrating brief content on stress management (e.g., the concept of growth mindset, normalizing that it's okay to struggle and seek help, tips on how to reframe challenges) could empower students. Some innovative approaches include mindfulness exercises tailored for students, or apps that check in on student well-being. Since our results show poor coping can lead to dropout, equipping students with even basic cognitive-behavioral tools (like how to challenge negative thoughts, how to break large tasks into smaller ones, etc.) is a preventative strategy. Academic advisors and counselors also play a role: reaching out to students who seem distressed and directing them to resources can prevent silent suffering that ends in withdrawal. Flexible Policies for Genuine Crises : Institutions should consider policies that allow temporary breaks or incompletes in courses when students face significant life disruptions. If a student knows they can take a short hiatus without penalty and resume the course, they might be less likely to fully drop out. This could be formalized as an "intermittent study plan" for those who need it. Our external stressor cases suggest that rigidity can turn a solvable interruption into a permanent dropout. Flexibility, coupled with support to get back on track, may convert some dropouts into persisters. 6.3 Limitations It is important to acknowledge the limitations of our study. First, the sample size and context: we had 20 students, mostly from a single institution/program. While QCA does not require large N and is suitable for small samples, the generalizability of specific configurations may be limited. Different contexts (e.g., K-12 online learning, corporate e-learning) might have other salient factors. Our study was also cross-sectional/retrospective in nature – we relied on students recalling their experiences, which may introduce recall bias. Additionally, while we tried to capture the temporal aspect (what led to what), fsQCA as applied here is not a dynamic analysis; it treats the presence of conditions and outcome in a static way. Dropout is a process that unfolds, and our representation simplifies that timeline. Another limitation is measurement: calibrating qualitative data into fuzzy sets involves subjective judgment. We mitigated this with careful definitions and double coding, but there is always some imprecision. Also, some constructs like "poor emotion regulation" were inferred rather than directly measured with a psychometric instrument, which could be improved in future work. 6.4 Conclusion Overall, our study contributes a more nuanced narrative to the story of online student dropout. It confirms that negative emotions are not just afterthoughts but are central to the dropout process, and it specifies how those emotions come about through particular combinations of stressors and lacks. It also confirms recent research that no one-size-fits-all explanation exists for dropout; instead, educators should envision multiple “persona” of at-risk students (the overwhelmed one, the isolated one, the un-coping one, etc.), each needing a tailored response. This configurational understanding encourages a shift from trying to single out the cause of dropout, to designing ecosystems that minimize multiple risks simultaneously and bolster multiple forms of support. Abbreviations fsQCA Fuzzy–set Qualitative Comparative Analysis ERT Emotional Regulation Theory COT Cognitive Overload Theory MOOCs Massive Open Online Courses LMS Learning Management System IDB Intermittent Discontinuance Behavior P E Fit–Person–Environment Fit Declarations Ethics approval and consent to participate This study was conducted in accordance with the Declaration of Helsinki, ensuring compliance with all ethical standards pertaining to research involving human participants. Informed consent was obtained from all participants prior to their involvement in the study. Ethical approval was granted by the Ethics Committee of Tongling Polytechnic (Approval No. 5). Consent for publication Not applicable. Competing Interests The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Funding This study was also sponsored by the following: 1. Middle-aged and young teachers training Action: cultivating outstanding young teachers’ program of Anhui Province (Grant No.: YQYB2023153). 2. Advanced teacher program of Tongling Polytechnic (Grant Name: Yao Yao) . 3. A Study on the Articulation Teaching of Specialized English for Cross-border E-commerce between Secondary Vocational and Higher Vocational Education (Grant No: tlpt2025jyzd06) Author Contribution Yao Yao: Conceptualization; Design; Drafting manuscript; Mingda Wang: Critical version of the manuscript; Supervision. Yang Zhang: Supervision. Acknowledgements This work was supported by Tongling Polytechnic, China and Universiti Kebangsaan Malaysia, Malaysia. Availability of data and materials The datasets generated and analyzed during the current study are openly available in figshare at DOI: 10.6084/m9.figshare.29517794 . References Abdelhalim SM. From Traditional Writing to Digital Multimodal Composing: Promoting High School EFL Students’ Writing Self-Regulation and Self-Efficacy. Comput Assist Lang Learn. 2024;55(1):1–30. 10.1080/09588221.2024.2322148 . Ahmed B, Yaru H, Rafiq M, Im C. Escaping the Boredom Trap: Enhancing Usability With 3D Games in Monotonous Subjects. IEEE Access. 2024;12:180361–84. 10.1109/ACCESS.2024.3434673 . Álvarez I, Manero B, Romero-Hernandez A, Cardenas M, Isabel, Masó. Virtual Reality Platform for Teacher Training on Classroom Climate Management: Evaluating User Acceptance. Virtual Reality. 2024;28. 10.1007/s10055-024-00973-6 . Ovidiu Ioan M. To Be (Online) or Not to Be? The Antecedents of Online Study Propensity and e-Learning-Dependent Dropout Intention in Higher Education. Technol Forecast Soc Chang. 2024;207:123566. 10.1016/j.techfore.2024.123566 . Bahtilla M. 2024. ‘Supervisory Feedback: Supervisors’ Reasons for Not Giving Timely Feedback’. Innovations in Education and Teaching International . Baltà-Salvador Rosó, Olmedo-Torre N, Peña M, Ana-Inés R-D. Academic and Emotional Effects of Online Learning during the COVID-19 Pandemic on Engineering Students. Educ Inform Technol. 2021;26(6):7407–34. 10.1007/s10639-021-10593-1 . Beard KS. Flow under Threat: Stereotype Threat and Flow Theory Integration in Support of Marginalized Student Academic Engagement and Psychological Well-Being. J Educ Psychol. 2025. 10.1037/edu0000951 . Behr A, Giese M, Teguim H, Theune K. Dropping out of University: A Literature Review. Rev Educ. 2020;8. 10.1002/rev3.3202 . Bekker CI, Rothmann S, and Magdalena M. Kloppers. The Happy Learner: Effects of Academic Boredom, Burnout, and Engagement. Front Psychol. 2023;13. 10.3389/fpsyg.2022.974486 . Bergdahl N. Engagement and Disengagement in Online Learning. Comput Educ. 2022;188:104561. 10.1016/j.compedu.2022.104561 . Boekaerts M. Emotions, Emotion Regulation, and Self-Regulation of Learning: Center for the Study of Learning and Instruction, Leiden University, The Netherlands, and KU Leuven’. in Handbook of Self-Regulation of Learning and Performance . Routledge; 2011. Chang C-C, Szu-Ting Y. Interactive Effects of Scaffolding Digital Game-Based Learning and Cognitive Style on Adult Learners’ Emotion, Cognitive Load and Learning Performance. Int J Educational Technol High Educ. 2023;20(1):16. 10.1186/s41239-023-00385-7 . Cloude EB, Wortha F, Dever DA. and Roger Azevedo. 2021. ‘Negative Emotional Dynamics Shape Cognition and Performance with MetaTutor: Toward Building Affect-Aware Systems’. Pp. 1–8 in 2021 9th International Conference on Affective Computing and Intelligent Interaction (ACII) . New York: Ieee. Cobo-Rendón R, Hojman V, García-Álvarez D, Ramon Cobo R. Academic Emotions, College Adjustment, and Dropout Intention in University Students. Front Educ. 2023;8. 10.3389/feduc.2023.1303765 . Delahunty J, Verenikina I. and Pauline Jones. 2014. ‘Socio-Emotional Connections: Identity, Belonging and Learning in Online Interactions. A Literature Review’. Technology, Pedagogy and Education . Derakhshan A, Kruk M, Mehdizadeh M, Mirosław P. Boredom in Online Classes in the Iranian EFL Context: Sources and Solutions. System. 2021;101:102556. 10.1016/j.system.2021.102556 . Eberle J, and Joyce Hobrecht. The Lonely Struggle with Autonomy: A Case Study of First-Year University Students’ Experiences during Emergency Online Teaching. Comput Hum Behav. 2021;121:106804. 10.1016/j.chb.2021.106804 . Eden AL, Johnson BK, Reinecke L, Grady SM. Media for Coping During COVID-19 Social Distancing: Stress, Anxiety, and Psychological Well-Being. Front Psychol. 2020;11. 10.3389/fpsyg.2020.577639 . Eres R, Lim MH, Lanham S, Jillard C, and Glen Bates. 2021. ‘LonelinessEmotion Regulation: Implications of Having Social Anxiety Disorder’. Australian J Psychol. Gross JJ. Emotion Regulation: Past, Present, Future. Cognition Emot. 1999. 10.1080/026999399379186 . He Y, Ding M, Liu M, Wu L, Liu Z, Guo H, Wang H. Teaching Reform and Practice Based on Four Dimensions and One Penetration for Sensing and Detection Technology. Int Rev Res Open Distrib Learn. 2024;25:66–86. 10.19173/irrodl.v25i4.7689 . Heng R, Cheng Q, Pu L, Tang Q. Pre-Service EFL Teachers’ Emotional Experiences, Emotion Regulation Strategies and Influencing Factors in Research: A Chinese Case Study. Heliyon. 2024;10(17). 10.1016/j.heliyon.2024.e36843 . Hongxia, Li, and Juan Yang. Managing Online Learning Burnout via Investigating the Role of Loneliness during COVID-19. BMC Psychol. 2025;13(1):151. 10.1186/s40359-025-02419-3 . Jimenez S, Salameh ME, Arguedas, Salvador Pons B. 2018. ‘Beyond the Notion of Periphery An Account of Polyfunctional Discourse Markers within the Val.Es.Co. Model of Discourse Segmentation’ edited by K. Beeching, C. Ghezzi, and P. Molinelli. POSITIONING THE SELF AND OTHERS: LINGUISTIC PERSPECTIVES 292:105–25. Kelly LA. 2023. ‘Using Engagement to Interrupt Mind-Wandering and Encourage Focus During Video-Recorded Lectures in the Online Learning Environment’. Ed.D. Koundal N, Abdalhadi A, Al-Quraishi MS, Elamvazuthi I, Moosavi MS, Guillet C, FRédéric, Merienne, Saad NM. 2024. ‘Effect of Interruptions and Cognitive Demand on Mental Workload: A Critical Review’. IEEE Access 12:54405–25. 10.1109/ACCESS.2024.3383142 Li D. The Shift to Online Classes during the COVID-19 Pandemic: Benefits, Challenges, and Required Improvements from the Students’ Perspective. Electron J E-Learning. 2022;20(1):1–18. 10.34190/ejel.20.1.2106 . Liu X, and Yijia Liu. Developing and Validating a Scale of Artificial Intelligence Anxiety Among Chinese EFL Teachers. Eur J Educ. 2025;60:e12902. 10.1111/ejed.12902 . Lv H, Low J, Tan Siow-kian, Tang L, Li X. Factors Affecting Medical Students’ Intention to Use Rain Classroom: A Cross-Sectional Survey. BMC Med Educ. 2024;24. 10.1186/s12909-024-05037-6 . MacIntyre PD, Gregersen T, Mercer S. Language Teachers’ Coping Strategies during the Covid-19 Conversion to Online Teaching: Correlations with Stress, Wellbeing and Negative Emotions. System. 2020;94:1–13. 10.1016/j.system.2020.102352 . McRae K. Cognitive Emotion Regulation: A Review of Theory and Scientific Findings. Curr Opin Behav Sci. 2016;10:119–24. 10.1016/j.cobeha.2016.06.004 . Miranda JO, Olga Angelinetta P, Tulabut, Reymond Neal C, Cruz. The PERMA Model of Well-Being and Student Engagement amid Covid-19: A Two-Wave Study among a Sample of Filipino University Students. Psychol Sch. 2024;61(11):35–51. 10.1002/pits.23286 . Peng P, Chen S, Hao Y, He L, Wang Q, Zhou Y, Tang Y-Y, Yang WF, Wu Q, Tieqiao Liu. Network of Burnout, Depression, Anxiety, and Dropout Intention in Medical Undergraduates. Int J Soc Psychiatry. 2023. 10.1177/00207640231166629 . Preston T, Carr DC, Hajcak G, Sheffler J, Natalie Sachs-Ericsson. Cognitive Reappraisal, Emotional Suppression, and Depressive and Anxiety Symptoms in Later Life: The Moderating Role of Gender. Aging & Mental Health; 2022. Ragin CC. Redesigning Social Inquiry: Fuzzy Sets and Beyond. University of Chicago Press; 2009. Rahmani A, Mohammad W, Groot, and Hamed Rahmani. Dropout in Online Higher Education: A Systematic Literature Review. Int J Educational Technol High Educ. 2024;21(1):19. 10.1186/s41239-024-00450-9 . Respondek L, Seufert T, Stupnisky R, Nett UE. Perceived Academic Control and Academic Emotions Predict Undergraduate University Student Success: Examining Effects on Dropout Intention and Achievement. Front Psychol. 2017;8. 10.3389/fpsyg.2017.00243 . Roopashree M, Praveen Kumar, Pavanalaxmi NS, Prameela, Mehnaz Fathima. ‘Multimedia Data Mod Education’. 2024. 10.1002/9781119786443.ch9 . Sanborn F. A Cognitive Psychology of Mass Communication. 8th ed. New York: Routledge; 2022. Shi C, Yu L, Wang N, Cheng B, and Xiongfei Cao. Effects of Social Media Overload on Academic Performance: A Stressor–Strain–Outcome Perspective. Asian J Communication. 2020;30:1–19. 10.1080/01292986.2020.1748073 . Solhi M. The Impact of EFL Learners’ Negative Emotional Orientations on (Un)Willingness to Communicate in In-Person and Online L2 Learning Contexts. J Psycholinguist Res. 2024;53(2). 10.1007/s10936-024-10071-y . Tian Y, Chan TJ, Liew TW, Chen MH, Huan Na Liu. 2025. ‘Overwhelmed Online: Investigating Perceived Overload Effects on Social Media Cognitive Fatigue via Stressor-Strain-Outcome Model’. Library Hi Tech ahead-of -print(ahead-of-print) . 10.1108/LHT-03-2024-0145 Tzafilkou K, Perifanou M, Economides AA. Negative Emotions, Cognitive Load, Acceptance, and Self-Perceived Learning Outcome in Emergency Remote Education during COVID-19. Educ Inform Technol. 2021;26(6):7497–521. 10.1007/s10639-021-10604-1 . Tze VMC, Lia M, Daniels, Klassen RM. Evaluating the Relationship Between Boredom and Academic Outcomes: A Meta-Analysis. Educational Psychol Rev. 2016;28(1):119–44. 10.1007/s10648-015-9301-y . Wu L, and Yingling Liu. Depression, Anxiety, and Stress among Vocational College Students during the Initial Stage of Post-Epidemic Era: A Cross-Sectional Study. Medicine. 2024;103(36):1–7. 10.1097/MD.0000000000039519 . Xu C, Xu Y, Xu S, Zhang Q, Liu X, Shao Y, Xu X, Peng L, Li M. Cognitive Reappraisal and the Association Between Perceived Stress and Anxiety Symptoms in COVID-19 Isolated People. Front Psychiatry. 2020;11. 10.3389/fpsyt.2020.00858 . Xue C. Mitigating EFL Students’ Academic Disengagement: The Role of Teachers’ Compassion and Mindfulness in China. HELIYON. 2023;9(2):e13150. 10.1016/j.heliyon.2023.e13150 . Yusof R, Harith NHM, Lokman A, Batau MFA, Zain ZM, Noor Hanim Rahmat. A Study of Perception on Students’ Motivation, Burnout and Reasons for Dropout. Int J Acad Res Bus Social Sci. 2023;13(7):403–32. 10.6007/IJARBSS/v13-i7/17187 . Zhang, Li, and Yunjie Chen. A Blended Learning Model Supported by MOOC/SPOC, Zoom, and Canvas in a Project-Based Academic Writing Course. Research-publishing.net; 2021. Zhao T, Fu Z, Lian X, Ye L, Huang W. Exploring Emotion Regulation and Perceived Control as Antecedents of Anxiety and Its Consequences During Covid-19 Full Remote Learning. Front Psychol. 2021;12. 10.3389/fpsyg.2021.675910 . Additional Declarations No competing interests reported. Supplementary Files Interviewguidance.docx questionnairedata.xlsx interviewdata.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 13 Feb, 2026 Reviewers agreed at journal 04 Sep, 2025 Reviewers agreed at journal 04 Sep, 2025 Reviewers invited by journal 28 Aug, 2025 Editor assigned by journal 13 Aug, 2025 Editor invited by journal 22 Jul, 2025 Submission checks completed at journal 13 Jul, 2025 First submitted to journal 13 Jul, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6971678","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":508030969,"identity":"6450f3f0-c529-4560-bdb1-d5ceec9eaaff","order_by":0,"name":"Yao Yao","email":"","orcid":"","institution":"Tongling Polytechnic","correspondingAuthor":false,"prefix":"","firstName":"Yao","middleName":"","lastName":"Yao","suffix":""},{"id":508030970,"identity":"65b6476a-5d96-4a00-a9a8-1c91dc4cc453","order_by":1,"name":"Mingda Wang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA0UlEQVRIie3RMQqDMBTG8RcCdtHOKYV4goIlUHDxLIpDF4WOjoqQLB7AwVsUXJvuOYBjXTo56Alau7WT6VZo/vsP3uMDMJl+MMvtxTRmD4pKLPXIGmKOicIMCx7qEQpHDg7HkaiUp3kYXPmNKIs5dTJ1AwR0ly8RVAjvlNl0U6dnv4GYHeQSwSgnRBG2r9N2a4OM2kVizcThXnTpkrsmsRGfSRgVlbI0CUElI0oyJDjzG0/jF7de9f2YydeUfTdkAV0kHxFbc5p38q0wmUymv+gJ2npBuz/lV1kAAAAASUVORK5CYII=","orcid":"","institution":"Universiti Kebangsaan Malaysia","correspondingAuthor":true,"prefix":"","firstName":"Mingda","middleName":"","lastName":"Wang","suffix":""},{"id":508030971,"identity":"86268ce6-fc55-488e-ba88-733f83e1d649","order_by":2,"name":"Yang Zhang","email":"","orcid":"","institution":"Tongling Polytechnic","correspondingAuthor":false,"prefix":"","firstName":"Yang","middleName":"","lastName":"Zhang","suffix":""}],"badges":[],"createdAt":"2025-06-25 07:24:13","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6971678/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6971678/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":90803006,"identity":"d17b530b-5271-4018-8eda-9fc3e1fe21f7","added_by":"auto","created_at":"2025-09-08 10:27:55","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":53259,"visible":true,"origin":"","legend":"\u003cp\u003eConceptual Diagram\u003c/p\u003e","description":"","filename":"Figure1ConceptualDigram.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6971678/v1/7f3cf33f73f069a9343b9410.jpg"},{"id":90803566,"identity":"7889971d-ec51-4158-b850-da71e12ed84e","added_by":"auto","created_at":"2025-09-08 10:35:56","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2202566,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6971678/v1/b721015d-18be-4438-b7de-c0a000edf0ae.pdf"},{"id":90801713,"identity":"2c9d85ce-573f-40e1-8ec5-8c2097673363","added_by":"auto","created_at":"2025-09-08 10:19:55","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":20207,"visible":true,"origin":"","legend":"","description":"","filename":"Interviewguidance.docx","url":"https://assets-eu.researchsquare.com/files/rs-6971678/v1/ed54eb64090ad4856d6f34a7.docx"},{"id":90803010,"identity":"e6aea97e-79a9-4ea5-9f1a-42f595e439ee","added_by":"auto","created_at":"2025-09-08 10:27:55","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":66208,"visible":true,"origin":"","legend":"","description":"","filename":"questionnairedata.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6971678/v1/e7f490fe32ca23db4b4f1f2b.xlsx"},{"id":90801725,"identity":"c32bedc4-90c9-4412-b448-fce1f9a1eed9","added_by":"auto","created_at":"2025-09-08 10:19:55","extension":"docx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":209338,"visible":true,"origin":"","legend":"","description":"","filename":"interviewdata.docx","url":"https://assets-eu.researchsquare.com/files/rs-6971678/v1/379691234ffc1249f25889cc.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Unravelling the Mechanisms of Student Negative Emotions in Technology-Mediated Intermittent Dropout: A Mixed-Methods fsQCA Study","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003eBy advocating with the benefits of on-line and technology assisted learning environments, they have expanded access to education, however they also face the challenges of student disengagement and dropout during the process. Unlike face to face teaching context, the online courses often report higher attrition rates. In a survey conducted with 300,000 students, it revealed that over 70% of them felt disengaged during the first learning period (Li \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The intermittent dropout, where students are temporarily disengaged or \u0026ldquo;drop out\u0026rdquo; during the virtual course learning, undermines the students' ability of consistent learning and engagement and raise the concern for educators and related stakeholders (Rahmani, Groot, and Rahmani 2024).\u003c/p\u003e\u003cp\u003eAntecedent literature has addressed the negative emotion affection as a pivotal role for students dropout behaviour. In contrast with face to face teaching, teachers always have the opportunities to observe student dropout behaviour with reminding and feedback to ease frustration or confusion immediately, online learners have to struggle with negative emotion in isolation (Eberle and Hobrecht 2021). Especially during the COVID-19 pandemic, studies claimed the high level anxiety, stress with other negative emotions among online environment learning students, which lead to excessive challenges mentally and chances of drop out (Balt\u0026agrave;-Salvador et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Li \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Based on the previous literature, the learners with high negative learning emotions \u0026ndash; anxiety, boredom, or depression- have more intention of withdraw or dropout from study comparing contradictory (Bekker, Rothmann, and Kloppers 2023; Respondek et al. \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Tze, Daniels, and Klassen 2016). However, studies on these emotional factors integration are still scant on the models of online persistence, while the mainstream interests are perspectives from cognitive, demographic, or infrastructural factors(Delahunty, Verenikina, and Jones 2014; Sanborn \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eVarious theoretical frameworks have been explored the mechanisms of emotional experiences might lead to learners dropout in digital assistant learning context. The Emotional Regulation Theory (ERT) stated mechanism of an individual manage and respond to their emotional states (Andrei Sitar-Tăut, Mican, and Ioan Moisescu 2024; Lv et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; McRae \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). As for learning context, the learner emotional regulation ability could affect its disengagement possibility (Boekaerts \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Liu and Liu \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Rahmani et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Studies have revealed that cognitive reappraisal could eliminate anxiety or depression level or strategy of suppression expression may lead to worsen stress (Preston et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Xu et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Actually, individual relying on suppression emotion experienced worst by negative factors and tend to report their feeling of more lonely, which in turn can heighten anxiety (Eres et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). This denotes that students cannot manage depression or anxiety successfully in online learning context, which may lead to burnout even learning dropout (Li and Yang 2025; Peng et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAnother relevant lens is the concept of information overload in digital environments. Online course are conducted with dense reading materials, multi-media visual and audio learning files, discussion activities, and countless notifications (Roopashree et al. \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Referring to Cognitive Overload Theory (COT), people attention is finite while overwhelming content and multi-dimensional requirements are likely trigger fatigue with distraction (Koundal et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Tian et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Recently, studies have shown that in the context of Massive Open Online Course (MOOCs) information overload of online learners experienced more anxiety and negative emotions maladapting their learning willingness and study adherence (Andrei Sitar-Tăut et al. 2024; Lv et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Rahmani et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Yusof et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Overloaded and overwhelmed, a student may feel compelled to step away from the course to recover. In fact, the high quality online course often features with dense reading material and high demanding interactions often witness higher students dropout rates (Andrei Sitar-Tăut et al. 2024; Yusof et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). This echoes with students burn out or stressed in online coursework reports (Shi et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAdditionally, according to the digital disengagement literature, social and motivational factors are related with online learning. Comparing to physical teaching classroom, the students of online learning are tend to experience more isolated or unsupported feeling(Balt\u0026agrave;-Salvador et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Eberle and Hobrecht 2021; Li and Yang 2025). Loneliness is not only a byproduct of online learning, but also a emotional situation to trigger learning burnout (Li and Yang 2025). Learning burnout as a negative emotion features exhaustion, cynicism, and could lead to student academic engagement (Peng et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Yusof et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Li and Yang (2025) claimed that loneliness could act as a maladaptive factor of burnout among other factors (Li and Yang 2025). Concurrently, more and more online learner report the feeling of boredom and interest decreasing caused by insufficient peer interaction or feedback from instructors (Jimenez, Arguedas, and Borderia 2018). By the disengagement and interest declining are accumulated or added over time, it eventually becomes disengagement with behaviors of intermittent dropout away from or avoiding online class platform (Andrei Sitar-Tăut et al. 2024; Bekker et al. 2023; Yusof et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Based on the literature, it can be founded that a combination of emotional stressors (anxiety, frustration, boredom) and contextual factors (isolation, overload, lack of support) are likely to integrated together leading to students learning intermittent dropout online (Yusof et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAlthough there are many studies conducted on the issue of intermittent dropout, the work on systematically study dropout in technology-medicated context is limited (Rahmani et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Previous studies typically investigated the individual predictors of online attrition (for instance, cognitive overload or loneliness or anxiety in isolation) by approaches of regression on correlated factors (Peng et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Rahmani et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Yusof et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). However, it is hard to justify the mechanism of student decision making from perspective of single factor, more recently there is a recognition that learners dropout is a combination of multiple converging factors and different students may leave for different combinations of reasons. Especially, in the environment of temporary disengagement and intermittent dropout, it is vital to explore how different factors jointly produce a tipping point where a student pauses or leaves a course.\u003c/p\u003e\u003cp\u003eTo answer this requirement, our study adopts a mixed method, configurational approach. We shall firstly investigate the student\u0026rsquo;s negative emotion mechanisms and their perceived triggers during episodes of disengagement. Then, using fuzzy-set Qualitative Comparative Analysis (fsQCA), we identify distinct configurations of factors that together lead to intermittent dropout. FsQCA is particularly suited to this investigation because it allows us to capture complex causality and equifinality (Ragin \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2009\u003c/span\u003e) \u0026ndash; the possibility that there are multiple, equally valid pathways to the outcome (dropout). We let patterns emerge from the data in a comparative case analysis; by doing so, this research strengthens the theoretical foundation of online student persistence by integrating emotional regulation perspectives with information overload and engagement frameworks. Practically, it yields insights into which combinations of issues (e.g. high workload and low support, or high anxiety and poor coping) are most dangerous, thereby informing more targeted interventions.\u003c/p\u003e\u003cp\u003eIn summary, this paper aims to unravel how negative emotions contribute to technology-mediated intermittent dropout through certain configurations of personal, emotional, and contextual factors. By following a clear mixed-methods design \u0026ndash; qualitative exploration followed by fsQCA \u0026ndash; we provide a well-rounded, evidence-based understanding of the dropout mechanisms. The article is structured as follows: first, we review relevant literature and theoretical frameworks on negative emotions, emotion regulation, overload, and disengagement in online learning. Next, we describe our methodology, including data collection, qualitative coding, calibration of variables, and fsQCA procedures. We then present the results, highlighting the key causal pathways identified. In the discussion, we interpret these findings in light of theory and prior research, discuss implications for educators and course designers, and acknowledge limitations and directions for future work.\u003c/p\u003e"},{"header":"2 Literature Review","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1 Negative Emotions and Online Student Dropout\u003c/h2\u003e\u003cp\u003eEmotions are critical to the learning experience, and negative emotions in particular can undermine motivation and persistence (Li and Yang 2025; Miranda, Tulabut, and Cruz 2024; Wu and Liu 2024). In traditional classrooms, emotions like anxiety or frustration are transiently observable and can be addressed through in-person support (\u0026Aacute;lvarez et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2024\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; He et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In online settings, however, such emotions often accumulate unnoticed until a breaking point. Recent studies have solidified the link between negative academic emotions and dropout intentions (Cobo-Rend\u0026oacute;n et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). For instance, Cobo-Rend\u0026oacute;n et al. (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) found that while positive emotions (enjoyment, hope) correlated with better adaptation and continued study, negative emotions were strong predictors of students\u0026rsquo; intentions to drop out of university. Notably, anxiety \u0026ndash; an emotion frequently reported in online courses \u0026ndash; has been specifically implicated in higher dropout rates (Peng et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). In a German university study, students with greater academic anxiety were significantly more likely to discontinue their studies (Cobo-Rend\u0026oacute;n et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). These findings align with the broader understanding that emotions can drive academic behavior: students overwhelmed by negative feelings often disengage as a form of escape or self-preservation (Beard \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Xue \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eCommon negative emotions in online learning include frustration, boredom, anxiety, and hopelessness (Cloude et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; MacIntyre, Gregersen, and Mercer \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Solhi M. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Frustration may arise from technical issues (e.g. platform glitches, poor internet connectivity) or confusing course materials (Cobo-Rend\u0026oacute;n et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Boredom often stems from a lack of interactivity or monotonous content delivery (Li and Yang 2025). Anxiety is frequently tied to high-stakes assessments, fear of falling behind, or simply the self-directed nature of e-learning which can cause students to worry if they are learning effectively (Liu and Liu \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Prolonged exposure to these aversive emotions can lead to what Li and Yang (2025) term learning burnout, defined as a state where students feel exhausted, detached, and ineffective in their learning (Li and Yang 2025). Burnout is essentially an amalgamation of chronic negative feelings, and it has been shown to seriously reduce student engagement in online instruction (Bekker et al. 2023). A burnt-out student often exhibits apathy (lack of interest in course content) and diminished participation, behaviors that are on the spectrum of disengagement and precursor to dropout (Peng et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIt is important to note that negative emotions do not operate in isolation; they are often symptomatic of underlying problems in the learning environment (Eberle and Hobrecht 2021; Eres et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Koundal et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). For example, a student might feel overwhelmed and anxious because the course workload is too high or instructions are unclear (He et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Tian et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Another might feel frustrated and helpless due to lack of timely feedback on assignments (Bahtilla \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). A sense of boredom or pointlessness might actually reflect insufficient interactivity or real-world relevance in the course (Ahmed et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Therefore, while we identify negative emotions as critical components in dropout cases, we must also examine their antecedents and context. By doing so, we can move beyond saying \"anxious students drop out\" to understand why the student became anxious to begin with (e.g. due to information overload or low self-efficacy) and how those factors combine to push the student toward disengaging.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e2.2 Emotion Regulation in Online Learning\u003c/h2\u003e\u003cp\u003eNot all students who experience negative emotions will drop out; the difference often lies in how they regulate and cope with those emotions (Behr et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Yusof et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Emotion regulation theory, as developed by Gross and others, provides a framework for understanding this process (Li and Yang 2025; Liu and Liu \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). According to Gross\u0026rsquo;s process model (Gross \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1999\u003c/span\u003e), individuals deploy strategies either before an emotion fully unfolds (antecedent-focused strategies) or after the emotion has already been generated (response-focused strategies). In an academic context, an example of an antecedent-focused strategy is cognitive reappraisal \u0026ndash; deliberately reframing a challenging situation in a more positive or neutral light (e.g., viewing a low quiz score as a learning opportunity rather than a catastrophe) (Zhao et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). A response-focused strategy example is expressive suppression \u0026ndash; concealing or inhibiting the outward signs of frustration or panic, without addressing the root cause of the emotion (Heng et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Each strategy impacts learning differently.\u003c/p\u003e\u003cp\u003eEmpirical evidence shows that adaptive emotion regulation is associated with better learning outcomes in online education. Zhao et al. (\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) demonstrated that students who habitually used cognitive reappraisal reported higher perceived control over their remote learning situation and ultimately learned more, whereas those who relied on suppression experienced heightened anxiety and learned less (Zhao et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Crucially, anxiety was found to be inversely related to learning in that study. This implies that failing to manage anxiety can directly impair performance and satisfaction, potentially prompting students to withdraw (Bekker et al. 2023; Rahmani et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In our context, this suggests that a student with good coping skills might endure a stressful online course without dropping out, whereas a student who cannot manage the stress might succumb to intermittent dropout.\u003c/p\u003e\u003cp\u003eMaladaptive regulation can also accumulate feelings of isolation and distress (Heng et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). As noted earlier, suppressing emotions (a common reaction for students hesitant to seek help) can lead to a buildup of negative affect (Eden et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Research during the pandemic observed that students who suppressed their worries were more likely to feel lonely and emotionally overwhelmed, compared to those who expressed and addressed their feelings (Li and Yang 2025; Preston et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Loneliness and anxiety reinforced each other in a vicious cycle. Ultimately, such students may disengage as a coping mechanism, essentially removing themselves from the source of stress (the course) because they see no other way to obtain relief (Xue \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eLiterature on self-regulated learning also intersects here. Emotion regulation is one facet of the broader ability of students to manage their learning process (Eres et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Gross \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1999\u003c/span\u003e). High-performing online learners often exhibit self-regulation skills: they set goals, manage time effectively, seek help proactively, and keep their motivation up (Chang and Yang 2023). These behaviors help deepen negative emotions (for instance, good time management can prevent last-minute panic). Conversely, students with poor self-regulation may quickly feel overwhelmed, which can spiral into frustration or despair (Abdelhalim \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In our study, we consider emotion regulation capacity as a potential condition influencing dropout. We posit that how a student handles negative feelings (by reappraising challenges or by internalizing distress) could be a differentiator in whether they drop out intermittently when faced with difficulties.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e2.3 Information Overload and Digital Disengagement\u003c/h2\u003e\u003cp\u003eThe design and delivery of online courses can carelessly contribute to negative emotions through information overload (Tzafilkou, Perifanou, and Economides \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Online learning requires students to process large amounts of digital information daily \u0026ndash; readings, videos, discussion forums, announcements, and more (35). When this volume of content exceeds a student's cognitive processing capacity or available study time, the result is often stress, confusion, and diminished learning (Li and Yang 2025). Chen et al. (2024) investigated high-quality MOOCs and found an intriguing paradox: courses with richer content and more interactive features (generally markers of quality) saw higher dropout rates, presumably because excessive content and interactions overwhelmed learners (Zhang and Chen 2021). They note that attention is a limited resource and scattering it across too many materials or activities can hinder learning outcomes. In practical terms, a student might start skipping videos or ignoring discussion boards to cope with overload, which then snowballs into disengagement (Kelly \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Stress and perceived difficulty from overload can manifest as anxiety and frustration, emotions strongly linked to dropout as discussed earlier.\u003c/p\u003e\u003cp\u003eDigital disengagement is a broader concept capturing the ways learners withdraw from online platforms, whether temporarily or permanently (Bergdahl \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Causes of disengagement are multifaceted. A systematic review by Rahmani et al. (\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) highlights factors such as screen fatigue, where prolonged time in front of a computer leads to exhaustion and loss of focus, and feelings of isolation, where students miss the social dimension of learning (Rahmani et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The rapid transition to online learning in the pandemic underscored these issues: many students reported boredom with online lectures, lack of motivation, and difficulty concentrating amid home distractions (Derakhshan et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Additionally, lack of prompt support (unable to get immediate clarification from an instructor) often left students feeling lost (Beard \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). These conditions contribute to what some authors call \"digital fatigue\" or \"e-learning fatigue,\" essentially a state of weariness and disillusionment with online study.\u003c/p\u003e\u003cp\u003eWhen students disengage digitally, it can take forms such as: not logging into the Learning Management System (LMS) for days or weeks, \u0026ldquo;ghosting\u0026rdquo; group projects, or ceasing to submit assignments \u0026ndash; all without formally un-enrolling from the course (Bergdahl \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). This intermittent dropout pattern can be seen as a coping response; students step away hoping to return once stress is lower or circumstances improve. Unfortunately, re-engaging after a long break is difficult, and many never fully catch up, leading to eventual failure or withdrawal. Previous research indicates that once disengaged, students face a high barrier to re-entry, especially if the course has moved on without them (Xue \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). For example, if 75% of students disengage early as noted in the German survey, only a fraction of those are likely to return and complete the course. Thus, preventing disengagement in the first place is critical.\u003c/p\u003e\u003cp\u003eIn summary, information overload and related design issues in online courses can evoke the very negative emotions that drive students away (Shi et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Tian et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Coupled with insufficient support or social interaction, these factors create a trigger for digital disengagement (Xue \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The literature underscores an important point: any successful intervention to reduce online dropout must address not just technical or academic preparation, but also the emotional and cognitive load placed on students. Reducing unnecessary complexity, balancing content quantity, fostering community, and providing timely help are all recommended practices to mitigate overload (Tian et al. \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Our study builds on this literature by examining how overload, isolation, and other factors combine with personal emotions in actual student dropout cases.\u003c/p\u003e\u003c/div\u003e"},{"header":"3 Conceptual Framework and Research Questions","content":"\u003cp\u003eBringing together the above strands, we conceptualize intermittent dropout in technology-mediated learning as an outcome emerging from multiple interacting factors. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e illustrates the proposed framework. In brief, environmental stressors (like heavy workload, confusing UI, or technical problems) can lead to information overload, which in turn provokes negative emotional responses (e.g., anxiety, frustration). Simultaneously, social/contextual deficiencies (like lack of instructor feedback or peer interaction) contribute to feelings of isolation or boredom, another set of negative emotions. The student's personal characteristics (such as emotion regulation ability, self-efficacy, time management skills) moderate how these experiences are internalized \u0026ndash; a resilient student might cope, whereas a vulnerable student might accumulate stress. When certain combinations of these conditions occur (e.g., high overload, low support, poor coping), the student crosses a threshold and disengages temporarily (an intermittent dropout event).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eWhat remains unclear, and forms the gap we address, is the exact combination of factors that lead to such dropout events. Traditional quantitative methods might tell us that \"overload increases odds of dropout by X%\" or \"anxiety correlates with dropout intentions,\" but they do not illuminate how factors cluster in individual cases (Rahmani et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Yusof et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). By using fsQCA, we approach this as a configurational problem: identifying sets of conditions that are sufficient to produce the outcome. This approach aligns with recent calls to use configurational comparative methods in education research to capture complex causality. While a few studies have applied QCA to topics like e-learning extension or student satisfaction, to our knowledge none have focused on emotional mechanisms in dropout behavior.\u003c/p\u003e\u003cp\u003eGiven the gap, our study raise questions:\u003c/p\u003e\u003cp\u003e\u003col\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eWhat configurations of negative emotions, personal factors, and contextual conditions lead to technology-mediated intermittent dropout?\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eWhat is the mechanism of student negative emotions in technology-mediated intermittent dropout?\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003c/ol\u003e\u003c/p\u003e\u003cp\u003eWe expect, based on the review above, that conditions such as high perceived workload, feelings of isolation, strong negative emotions (anxiety, frustration), and poor coping ability will feature prominently in the dropout pathways. Conversely, we anticipate that absence of these issues or presence of protective factors (e.g., high support, effective emotion regulation) will distinguish those who persist. These expectations inform our analysis; the fsQCA will reveal which combinations actually mattered for our participants.\u003c/p\u003e\u003cp\u003eBy addressing this research objective, our study contributes to theory by integrating emotional and cognitive perspectives on dropout, and to practice by identifying leverage points (which factor combinations to target) to reduce online student attrition. The next section details how we designed the study and carried out the mixed-methods investigation to achieve these aims.\u003c/p\u003e"},{"header":"4 Methodology","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e4.1 Research Design\u003c/h2\u003e\u003cp\u003e We employed a mixed-methods sequential exploratory design consisting of a qualitative phase followed by a quantitative fsQCA phase. The rationale for this design was to first gain an in-depth understanding of students' emotional experiences and the context of their intermittent dropout (through qualitative inquiry), and then to systematically analyze patterns across cases using fsQCA. This approach allowed us to ground the quantitative analysis in real-world insights and to ensure that the conditions fed into the fsQCA model were relevant and evidence-based. The design can be summarized in two broad stages:\u003c/p\u003e\u003cp\u003e\u003col\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eQualitative Exploration: We conducted semi-structured interviews with students who had experienced intermittent dropout in an online learning context. The goal was to uncover which factors (emotional, academic, environmental, etc.) they perceived as contributing to their disengagement episodes, and how those factors interrelated. We also probed how students responded to challenges (coping strategies) to assess their emotion regulation in context.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eConfigurational Analysis (fsQCA): Based on the qualitative results, we identified key conditions and calibrated these as fuzzy-set variables. Each student's case (from the interviews, and supplemented by additional data where available) was then represented as a case in a comparative analysis. Using fsQCA software, we examined which combinations of conditions were present in cases where dropout occurred, thereby identifying sufficient causal recipes for intermittent dropout.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003c/ol\u003e\u003c/p\u003e\u003cp\u003eThis design aligns with fsQCA's strengths in handling complex causality and small-to-medium sample sizes, while the qualitative component enhances the study's credibility and contextual richness. By clearly explaining each step \u0026ndash; from qualitative coding to variable calibration to solution interpretation \u0026ndash; we heed the reviewer's call for a transparent research process.\u003c/p\u003e\u003cp\u003eEthical approval was obtained from the Tongling Polytechnic Academic Committee prior to data collection. All participants gave informed consent, and their identities were kept confidential (pseudonyms or ID codes were used in lieu of real names in transcripts and analysis). Participation was voluntary, and students were assured that declining or withdrawing would have no effect on their academic standing. These measures ensured an ethical and open environment for participants to share their experiences of dropout.\u003c/p\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e4.2 Qualitative Phase: Interviews and Coding\u003c/h2\u003e\u003cp\u003e\u003cb\u003eParticipants\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe qualitative phase involved N\u0026thinsp;=\u0026thinsp;20 participants recruited from a large online degree program at a university (or a MOOC platform, etc.). We used purposive sampling to select students who had shown evidence of intermittent dropout, defined operationally as having temporarily stopped engaging with an online course for a significant period (e.g., missing several weeks of course participation or assignments) before either returning or eventually withdrawing. The sample (detailed in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) included a mix of genders, and majors to capture diverse perspectives. The mean age was roughly 20 (with a range from late teens to mid-career adults, reflecting typical online learner demographics). We stopped recruitment at 20 interviews as we reached thematic saturation (no fundamentally new themes were emerging in later interviews).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eParticipant Information Statistics Table\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"3\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCategory\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDescription\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003ePercentage/Number\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eGender\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eFemale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e49.1% (n\u0026thinsp;=\u0026thinsp;146)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e50.9% (n\u0026thinsp;=\u0026thinsp;150)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge Distribution\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e18\u0026ndash;24 years old\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e100% (n\u0026thinsp;=\u0026thinsp;296)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEducational Background\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCollege students\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e100% (n\u0026thinsp;=\u0026thinsp;296)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e\u003cp\u003eFrequency of Educational Technology Use\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDaily use\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e70 participants (23.64%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSeveral times a week\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e186 participants (62.41%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eOnce a week or less\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e40 participants (13.51%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"4\" rowspan=\"5\"\u003e\u003cp\u003eExperience with IDB\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eIntermittent Discontinuance Behavior\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e214 participants (72.29%)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAnxiety\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e83.6%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDissatisfaction\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e76.9%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDepression\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e68.2%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWorries\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e71.5%\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eData Collection\u003c/b\u003e\u003c/p\u003e\u003cp\u003eWe developed a semi-structured interview guide with open-ended questions covering areas such as: students\u0026rsquo; overall online learning experience, specific instances of disengagement (e.g., \"Can you describe a time when you stopped participating in an online course? What was happening at that time?\"), emotions felt during that period (prompting for feelings like stress, frustration, boredom, etc.), perceived causes or triggers of those emotions, support or lack thereof from instructors/peers, and any strategies they used to cope or re-engage. Follow-up questions probed deeper into key statements (for example, if a student said \"I was completely overwhelmed,\" we asked what contributed to that feeling). Interviews were conducted via video conferencing and lasted approximately 25\u0026ndash;40 minutes each. All interviews were audio-recorded (with permission) and transcribed verbatim for analysis.\u003c/p\u003e\u003cp\u003e\u003cb\u003eQualitative Analysis\u003c/b\u003e\u003c/p\u003e\u003cp\u003eWe used a thematic analysis approach (following Braun \u0026amp; Clarke's guidelines) to code and interpret the interview data. Initially, two researchers independently read through a subset of transcripts to familiarize themselves with the content and note preliminary ideas. Next, they collaboratively developed a coding scheme that balanced deductive codes (derived from our conceptual framework, e.g., \"technical difficulties\", \"workload\", \"anxiety\", \"seeking help\") and inductive codes (new themes emerging from the data, e.g., \"family responsibilities\" as a reason to drop out, if it arose). Each transcript was coded line-by-line in qualitative analysis NVivo software. The two researchers double coded about 25% of the transcripts to ensure reliability, achieving an inter-coder agreement of ~\u0026thinsp;0.85 (Cohen\u0026rsquo;s kappa) on key thematic categories after resolving minor discrepancies through discussion.\u003c/p\u003e\u003cp\u003eThrough this coding process, we identified recurrent factors associated with the students\u0026rsquo; dropout episodes. These factors (which later informed our fsQCA conditions) included, for example:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eOverwhelming Workload/Content: Many students described the course content as too dense or fast-paced, using phrases like \"information overload\" or \"swamped with work.\" This often led to feelings of anxiety and helplessness.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eTechnical or Logistical Issues: Some cited persistent technical breakdown, unreliable internet, or difficulty using the learning platform, which caused frustration and disrupted their learning routine.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eLack of Instructor Support: A common theme was that when students struggled, they did not get timely help or feedback. Emails or forum questions went unanswered, leaving them feeling isolated (\u003cem\u003e\"I was on my own\"\u003c/em\u003e).\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eLow Peer Interaction: Especially during pandemic remote learning, students missed having classmates to study with or simply to commiserate. The absence of a learning community contributed to boredom and a sense that \u003cem\u003e\u0026ldquo;no one would notice if I disappeared.\u0026rdquo;\u003c/em\u003e\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003ePersonal Life Stressors: External factors like work commitments, family responsibilities, or health issues sometimes precipitated disengagement. These were often accompanied by stress and guilt, and if the course was not flexible, the student would drop out to cope.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eNegative Emotional States: Across nearly all interviews, students explicitly mentioned emotions \u0026ndash; \u003cem\u003efrustration\u003c/em\u003e (\u0026ldquo;I got so frustrated that I gave up for a while\u0026rdquo;), \u003cem\u003eanxiety\u003c/em\u003e (\u0026ldquo;I was constantly anxious about falling behind\u0026rdquo;), \u003cem\u003eboredom\u003c/em\u003e (\u0026ldquo;The class was just so dull, I lost motivation\u0026rdquo;), and \u003cem\u003eshame\u003c/em\u003e or \u003cem\u003elow self-efficacy\u003c/em\u003e (\u0026ldquo;I felt I wasn\u0026rsquo;t smart enough to do the assignments\u0026rdquo;). These emotions were described both as outcomes of the above factors and as immediate triggers for stepping away.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eCoping and Regulation: We observed differences in how students handled their emotions. Some tried positive coping (contacting the instructor, taking a short break and coming back), whereas others resorted to avoidance (completely logging off, ignoring the problem). Notably, very few had formal strategies to regulate emotions; most actions were ad-hoc. Students who eventually returned to the course sometimes cited an external push (deadline pressure, a reminder from a friend, etc.) rather than intrinsic regulation.\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eThrough iterative discussion and concept mapping, we synthesized these findings into a set of candidate conditions to be used in the fsQCA. We ensured each condition was clearly defined and grounded in multiple interviewees\u0026rsquo; accounts (not a one-off mention). For example, \u003cem\u003e\u0026ldquo;Overload\u0026rdquo;\u003c/em\u003e was defined as the student feeling the course demands exceeded their capacity, evidenced by quotes about excessive content or time pressure; \u003cem\u003e\u0026ldquo;Lack of support\u0026rdquo;\u003c/em\u003e was defined by instances where students could not get help when needed; \u003cem\u003e\u0026ldquo;High negative emotion\u0026rdquo;\u003c/em\u003e was defined by intense feelings like severe anxiety or frustration that the student linked to their dropout decision. Importantly, we also noted cases of \u003cem\u003enon-dropout within our sample\u003c/em\u003e (some students described difficult situations which they managed to overcome and did \u003cem\u003enot\u003c/em\u003e drop out). These provided contrastive insights \u0026ndash; e.g., a student who struggled but persisted often had at least one mitigating factor (like strong self-motivation or external support).\u003c/p\u003e\u003cp\u003eThe outcome of the qualitative phase was thus a nuanced understanding of the mechanisms connecting various factors to negative emotions and dropout. We compiled a summary for each interviewee (case) detailing whether each of the identified factors was present or salient in their story. This case-wise data became the basis for calibrating fuzzy-set membership scores in the next phase.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e4.3 Calibration of Variables for fsQCA\u003c/h2\u003e\u003cp\u003eUsing the qualitative insights, we proceeded to operationalize the conditions and outcome for the fsQCA. FsQCA requires each case (here, each student) to have a score between 0 and 1 for each condition and the outcome, representing degree of membership in the set (e.g., the set of students experiencing overload, or the set of students who dropped out). We defined the following key outcome and conditions detailed in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e:\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eCalibration Table\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVariables\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eFully Membership (0.0)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003ePartial Membership (0.5)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eFully Non-Membership (1.0)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIntermittent Dropout\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eClear episode of disengagement\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eReduced activity, borderline case\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNon-dropout\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOverload\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eStrongly felt overload\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eModerate overload at times\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNever mentioned workload\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLack of support\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eFelt adequately support\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eMinimal or delayed support\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eExperienced major support gaps\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNegative Emotions\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNot report\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eModerate negative emotions present\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eStrong negative emotions described\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePoor Emotional regulation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eProactive and adaptive coping\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003ePartial\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNot cope or maladaptive\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIsolation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eFelt well-connected\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSome interaction insufficient\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eFelt very isolated\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eOutcome (Intermittent Dropout)\u003c/b\u003e: This was the target set we are trying to explain. We coded a case as 1 (full membership) in dropout if the student had a clear episode of disengagement (e.g., stopped participating for a significant period or officially withdrew) during the course in question. If a student did not disengage at all (continuous participation), they would be 0 (full non-membership). Partial membership (0.5) could be assigned in borderline cases \u0026ndash; for instance, if a student reduced activity for a while but did not completely stop (though we tried to keep outcome binary for interpretability). In our data, since we specifically sampled dropout cases, most cases were 1; however, we included a few non-dropout cases (students who nearly dropped but managed to continue) as counterexamples to enrich the analysis.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eCondition 1: Overload (OL)\u003c/b\u003e \u0026ndash; Reflecting \u003cb\u003einformation overload or overwhelming workload\u003c/b\u003e. Using interview evidence and any available performance data (like hours spent or number of assignments overlapping), we calibrated this as: 1 if the student strongly felt overloaded (explicit statements of being overwhelmed, consistently unable to keep up with content), 0.5 if moderate (some overload at times, but manageable or mixed experience), and 0 if overload was not a factor (student never mentioned workload as an issue or explicitly said workload was fine).\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eCondition 2: Lack of Support (LS)\u003c/b\u003e \u0026ndash; Capturing \u003cb\u003einsufficient instructor or institutional support\u003c/b\u003e. We gave 1 if the student experienced major support gaps (e.g., no response when seeking help, felt completely unsupported), 0.5 if support was minimal or delayed (e.g., help came but too late or only after repeated requests), and 0 if the student felt adequately supported (instructor was responsive, resources were sufficient).\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eCondition 3: Negative Emotions (NE)\u003c/b\u003e \u0026ndash; Indicating the presence of \u003cb\u003eintense negative emotional experience\u003c/b\u003e during the course. We focused on emotions like anxiety, frustration, and demotivation. 1 meant the student described strong negative emotions that significantly affected them (e.g., panic attacks, extreme frustration leading to anger or tears), 0.5 for moderate negative emotions (stress was present but perhaps not incapacitating, or fluctuated), and 0 if the student did not report notable negative feelings (this was rare in our dropout cases, more common in non-dropouts).\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eCondition 4: Poor Emotion Regulation (PR)\u003c/b\u003e \u0026ndash; Representing \u003cb\u003eineffective coping or regulation strategies\u003c/b\u003e. This was a more inferential measure based on interview responses to how they dealt with challenges. 1 was assigned if the student basically \u003cem\u003edid not cope\u003c/em\u003e or used maladaptive strategies (withdrew completely, ignored issues, self-blame without action), 0.5 if they made some effort but it was partial (e.g., took a break but maybe too late, or tried one strategy that didn\u0026rsquo;t fully work), and 0 if they demonstrated proactive and adaptive coping (sought help, adjusted study strategies, used positive reframing). We cross-validated this with their outcome \u0026ndash; often, dropouts had high PR (poor regulation) whereas those who navigated difficulties had low PR (good regulation).\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eCondition 5: Isolation (IS)\u003c/b\u003e \u0026ndash; Reflecting \u003cb\u003elow social presence or peer engagement\u003c/b\u003e. \u003cb\u003e1\u003c/b\u003e if the student felt very isolated (no peer interaction, explicitly complained of loneliness or lack of collaboration), \u003cb\u003e0.5\u003c/b\u003e if there was some interaction but insufficient (maybe they had group work but it was perfunctory, or they felt \u0026ldquo;disconnected\u0026rdquo; even though others were technically present), and \u003cb\u003e0\u003c/b\u003e if the student felt well-connected (had active communication with peers or instructors). This condition is tied to disengagement from the community aspect of learning.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cem\u003e(Additional conditions could include personal factors like self-efficacy or external pressures, if our qualitative data strongly indicated them. For brevity, we list five above, which were among the most recurrent.)\u003c/em\u003e\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eEach case (student) thus could be represented as a vector of these scores. We employed the direct method of calibration, where we set three qualitative anchors for each fuzzy set: one for full membership (criterion for being fully in the set), one for the crossover point (0.5, the point of maximum ambiguity where the case is neither in nor out), and one for full non-membership. These were based on our code definitions and, where possible, numeric indicators. For example, for \u003cem\u003eOverload\u003c/em\u003e, if a student explicitly said \"I had no trouble with the workload,\" that was evidence for a score near 0; if they said \"I was drowning in readings and assignments,\" that was evidence for a score of 1. Intermediate statements (or mixed signals) would lead to a 0.5. By anchoring the calibration to qualitative evidence (often using participant quotes as benchmarks), we ensured our fuzzy-set scores stayed true to the data (a process akin to \u003cem\u003e\u0026ldquo;anchored calibration\u0026rdquo;\u003c/em\u003e in QCA methodology). To enhance reliability, a second researcher reviewed the calibration of each case; disagreements (e.g., whether a certain description warranted a 0.5 or 1) were resolved through discussion, referring back to transcripts.\u003c/p\u003e\u003cp\u003eWe used fsQCA software (Version 3.0) to input the calibrated data for analysis. Before constructing the truth table, we checked for necessary conditions. A necessity analysis assesses if any single condition is present in \u003cem\u003eall\u003c/em\u003e or nearly all instances of the outcome (i.e., does any factor appear to be a prerequisite for dropout?). We found no condition with a consistency above the typical threshold of 0.90 for necessity \u0026ndash; in other words, no single factor was absolutely required for an intermittent dropout to occur. This reinforces the idea that different combinations can lead to dropout (equifinality), and it justified our focus on sufficiency analysis via the truth table.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e4.4 fsQCA Procedure and Solution Derivation\u003c/h2\u003e\u003cp\u003eWith calibrated data ready, we proceeded to the core fsQCA analysis:\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003e\u003cb\u003eTruth Table Construction\u003c/b\u003e\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe software generated a truth table, which lists all possible combinations of the conditions (2^5\u0026thinsp;=\u0026thinsp;32 possible rows, for five conditions) and indicates which combinations are present in our observed cases and whether they lead to the outcome. Given our sample size (~\u0026thinsp;20 cases), many logically possible combinations had no cases (so-called \u0026ldquo;logical remainders\u0026rdquo;). We set a frequency threshold of 1 case per combination to consider a configuration empirically relevant (since our sample is small, even a configuration with 1 case is of interest, but in larger samples one might set this higher). We set a consistency threshold of ~\u0026thinsp;0.80 for a combination to be considered sufficient for the outcome. Consistency here measures how uniformly cases with that combination exhibit the outcome \u0026ndash; values close to 1 indicate that whenever that combination is present, the outcome is almost always present as well. We examined the truth table rows exceeding 0.8 consistency and flagged them for inclusion in the logical minimization.\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eLogical Minimization: Using QCA\u0026rsquo;s minimization algorithm (Quine\u0026ndash;McCluskey), we derived simplified \u003cem\u003esolution terms\u003c/em\u003e that explain the outcome. We opted to derive the intermediate solution, which balances between the complex solution (no remainders used, very conservative) and the parsimonious solution (maximal use of remainders, risk of over-simplification). To produce an intermediate solution, one must specify directional expectations for how each condition relates to the outcome (i.e., whether the presence or absence of each condition is theorized to contribute to dropout). Based on literature and our qualitative insights, we assumed that the \u003cem\u003epresence\u003c/em\u003e of risk factors (overload, lack of support, strong negative emotions, poor regulation, isolation) would favor the occurrence of dropout, whereas their \u003cem\u003eabsence\u003c/em\u003e or opposites would favor persistence. These assumptions guided the inclusion of logical remainders that did not contradict our theoretical understanding. Importantly, we kept these expectations moderate. The software then provided the intermediate solution, consisting of one or more configurations (causal recipes) that are sufficient for the outcome.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eSolution Evaluation: We recorded the consistency and coverage of each configuration and of the overall solution. Consistency of each solution term indicates how reliably that combination leads to dropout (again we sought values well above 0.8). Raw coverage tells us the proportion of outcome cases explained by that term, and unique coverage indicates the proportion explained \u003cem\u003eonly\u003c/em\u003e by that term (i.e., cases not covered by other terms). We also looked at the solution consistency and coverage as a whole. A high solution consistency (e.g., \u0026gt;\u0026thinsp;0.8) means the set of identified configurations collectively are good predictors of dropout cases, and a decent solution coverage (say 0.5\u0026ndash;0.7) means they explain a substantial fraction of all dropout cases, recognizing that some cases might remain unaccounted for if they had very idiosyncratic causes.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eRobustness and Sensitivity Checks: To ensure our findings were not an artifact of specific parameter choices, we conducted several checks:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eWe raised the consistency threshold to 0.85 to see if the solutions changed; the core configurations remained the same, giving confidence in their robustness.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eWe tried calibrating the outcome as a fuzzy set (for those who partially disengaged) instead of a crisp set; this did not substantially alter which conditions appeared important.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eWe ran an analysis of the negation of the outcome (i.e., examining configurations for students who persisted without dropout) to see if it yielded simply the mirror image of dropout configs or something different. This revealed, interestingly, that persistence often required the absence of multiple risk factors simultaneously (no overload \u003cem\u003eand\u003c/em\u003e good support \u003cem\u003eand\u003c/em\u003e so on), reinforcing that dropout can be triggered by the presence of any one pathway of problems, whereas staying engaged may demand that several things go right at once.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eWe also tested removing one condition at a time to check if any single condition was overly influential (e.g., if removing \"Negative Emotions\" drastically drops solution consistency, it means most paths relied on that condition). We found that while some conditions (like NE) appeared in multiple solutions, the overall phenomenon could still be explained in their absence via other routes, confirming that multiple independent pathways exist.\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eThroughout the analysis, we prioritized a \u003cb\u003etransparent and data-driven approach\u003c/b\u003e. All analytic decisions (calibration thresholds, consistency cut-off, inclusion of remainders) were documented and can be traced back to either empirical evidence or theoretical rationale. By avoiding arbitrary or biased choices, we aimed to let the data \u0026ldquo;speak for itself,\u0026rdquo; fulfilling the review requirement to let conclusions emerge from the fsQCA rather than imposing preconceived hypotheses.\u003c/p\u003e\u003cp\u003eIn the next section, we present the results of the fsQCA, including the specific configurations identified as leading to intermittent dropout, illustrated with examples from the qualitative cases to give them concrete meaning.\u003c/p\u003e\u003c/div\u003e"},{"header":"5 Results","content":"\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003e5.1 Qualitative Insights: Factors Precipitating Dropout\u003c/h2\u003e\u003cp\u003eBefore delving into the fsQCA configurations, it is helpful to briefly summarize the qualitative insights that set the stage. As noted, the interviews revealed a constellation of factors that frequently co-occurred with students\u0026rsquo; decisions to temporarily drop out. To illustrate, we highlight a few anonymized cases:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eCase A (Overload \u0026amp; Anxiety): A first-year student described how a combination of \u003cem\u003efive challenging courses\u003c/em\u003e and \u003cem\u003econtinuous weekly assignments\u003c/em\u003e led to a crushing workload. She recounted, \u003cem\u003e\u0026ldquo;Every day there was something due. I couldn\u0026rsquo;t keep track.\u0026rdquo;\u003c/em\u003e This overload triggered intense anxiety and sleepless nights. Without an outlet or guidance on managing the load, she stopped logging into two of her courses for a month (an intermittent dropout) to focus on her mental health.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eCase B (Isolation \u0026amp; Boredom): Another student, enrolled in a fully online program during the pandemic, felt extreme isolation. He said, \u003cem\u003e\u0026ldquo;I never really got to know my classmates. Discussion boards felt like talking to a void.\u0026rdquo;\u003c/em\u003e Lacking social interaction, he lost interest and described the course as boring. Midway, he disengaged for several weeks, saying \u003cem\u003e\u0026ldquo;I figured it didn\u0026rsquo;t matter to anyone if I showed up or not.\u0026rdquo;\u003c/em\u003e\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eCase C (Tech Issues \u0026amp; Frustration): A working adult student faced recurrent technical problems \u0026ndash; the learning platform would crash during quizzes, and video lectures were often inaccessible on her network. These issues led to mounting frustration: \u003cem\u003e\u0026ldquo;I was spending more time troubleshooting than learning.\u0026rdquo;\u003c/em\u003e After failed attempts to get IT support, she temporarily dropped out in exasperation.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eCase D (Personal Stress \u0026amp; Lack of Support): A part-time student balancing family responsibilities had a medical emergency at home. She informed the instructor she would be offline for a week, but got no extension or meaningful help catching up. Feeling overwhelmed by personal stress and unsupported academically, she disengaged. She later said, \u003cem\u003e\u0026ldquo;If the professor had just reached out or given me some leeway, I might have continued.\u0026rdquo;\u003c/em\u003e\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eThese examples show how multiple factors converged in each dropout episode. Importantly, not every case had all factors present; rather, \u003cem\u003edifferent subsets of factors\u003c/em\u003e drove different students to disengage. This observation is what the fsQCA results formalize.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003e5.2 Configurations Leading to Intermittent Dropout (fsQCA Findings)\u003c/h2\u003e\u003cp\u003eThe fsQCA yielded several configurations (causal pathways) that were sufficient to produce the outcome of intermittent dropout. Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e below summarizes the simplified solution terms obtained, with each configuration indicating a combination of conditions. (Note: In a text format here, we describe them in prose, but one could imagine a table with conditions marked present (●), absent (\u0026otimes;), or irrelevant for each solution, alongside consistency and coverage metrics.)\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eConfiguration path leading to Intermittent Dropout\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eVariables\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003eSolution Cases\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003ePath 1\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003ePath 2\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003ePath 3\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003ePath 4\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOverload\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e●\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u0026otimes;\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u0026otimes;\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u0026otimes;\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLack of Support\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e●\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u0026otimes;\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u0026otimes;\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u0026otimes;\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNegative Emotions\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e●\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e●\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e●\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u0026otimes;\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIsolation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u0026otimes;\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e●\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u0026otimes;\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u0026otimes;\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePoor Emotion Regulation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u0026otimes;\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u0026otimes;\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e●\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u0026otimes;\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRaw Coverage\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.45\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.30\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.25\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUnique Coverage\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.013\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.011\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.06\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eConsistency\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.92\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.88\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.85\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSolution Coverage\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e0.849\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSolution Consistency\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\u003cp\u003e0.811\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eSolution 1: Overload\u003c/b\u003e \u003cb\u003eAND\u003c/b\u003e \u003cb\u003eLack of Support\u003c/b\u003e \u003cb\u003eAND\u003c/b\u003e \u003cb\u003eNegative Emotions\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThis configuration can be expressed logically as: OL * LS * NE (meaning overload, combined with lack of support, combined with presence of strong negative emotions). Consistency\u0026thinsp;=\u0026thinsp;0.92, Raw coverage\u0026thinsp;=\u0026thinsp;0.45. In plain terms, Solution 1 indicates that when students experience a heavy workload or information overload (OL\u0026thinsp;=\u0026thinsp;1) \u003cem\u003etogether with\u003c/em\u003e insufficient support from instructors (LS\u0026thinsp;=\u0026thinsp;1), and as a result they suffer intense negative emotions like anxiety or frustration (NE\u0026thinsp;=\u0026thinsp;1), they are very likely to drop out intermittently. Several cases aligned with this pattern. For example, in Case A above, we see exactly this: enormous content demands plus no felt support, yielding severe anxiety. Most students under this triple strain reached a breaking point. Notably, this path reflects a classic overload-burnout scenario: the environment overloads the student, the student doesn't get help, and their emotional state collapses, forcing disengagement.\u003c/p\u003e\u003cp\u003e\u003cb\u003eSolution 2: Isolation\u003c/b\u003e \u003cb\u003eAND\u003c/b\u003e \u003cb\u003eBoredom (or Lack of Interest)\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThis configuration, simplified, corresponds to IS * (lack of engagement). In terms of our conditions, that could be captured by IS\u0026thinsp;=\u0026thinsp;1 (isolation present) and perhaps also NE\u0026thinsp;=\u0026thinsp;1 if boredom is considered a negative emotion. In our formal model, boredom was one aspect of NE, but here it's specifically tied to isolation. Consistency\u0026thinsp;=\u0026thinsp;0.88, Raw coverage\u0026thinsp;=\u0026thinsp;0.30. Solution 2 suggests that social isolation, coupled with a resulting loss of interest or boredom, is another pathway to dropout. This was evident in Case B: feeling no social connection (IS) led to disengagement out of apathy. Interestingly, in some of these cases the workload was not necessarily high \u0026ndash; it was the lack of \u003cem\u003emeaningful interaction\u003c/em\u003e that made the course feel not worth pursuing. This configuration resonates with the concept that emotional disengagement (boredom, apathy) can be just as detrimental as cognitive overload. It also highlights the importance of social presence: in our data, no student who felt well-integrated socially dropped out purely from boredom, indicating isolation was a key ingredient.\u003c/p\u003e\u003cp\u003e\u003cb\u003eSolution 3: Poor Emotion Regulation\u003c/b\u003e \u003cb\u003eAND\u003c/b\u003e \u003cb\u003eHigh Negative Emotion (even if workload manageable)\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThis path can be denoted as PR * NE, with absence of OL (i.e., ~OL) in some instances. Consistency\u0026thinsp;=\u0026thinsp;0.85, Coverage\u0026thinsp;=\u0026thinsp;0.25. It points out cases where a student\u0026rsquo;s inability to cope (PR\u0026thinsp;=\u0026thinsp;1) with even moderate challenges led to runaway negative emotions (NE\u0026thinsp;=\u0026thinsp;1) and eventually dropout. In some instances here, the objective course demands were not extreme (so overload was absent or low), but the student\u0026rsquo;s personal coping threshold was low. For example, a student might have had perfectionist tendencies or high anxiety trait; a small setback (like one poor grade) spiraled into major self-doubt and panic because they didn\u0026rsquo;t utilize healthy coping strategies (no reappraisal, only rumination). These students often withdrew even though external conditions alone might not predict it. Solution 3 underscores the role of individual differences in emotion regulation \u0026ndash; even a well-designed course can lose students if they lack resilience or coping mechanisms. Conversely, it implies that building students\u0026rsquo; self-regulatory skills could prevent dropout in cases where academic factors are surmountable.\u003c/p\u003e\u003cp\u003e\u003cb\u003eSolution 4: External Stressor\u003c/b\u003e \u003cb\u003eAND\u003c/b\u003e \u003cb\u003eLack of Flexibility\u003c/b\u003e (a less frequent but present pathway):\u003c/p\u003e\u003cp\u003eAlthough not among the top three configurations by coverage, we observed a pattern where an external life stressor (like health or family issues, not formally one of our main conditions but qualitatively noted) combined with an inflexible course structure (could be seen as a special case of lack of support or a high workload that can't be adjusted) led to dropout. This is effectively a conjunctural cause outside the core model but worth mentioning. Consistency was high for these few cases. It reminds us that sometimes the trigger to disengage comes from outside academia, but whether it results in dropout depends on the course\u0026rsquo;s accommodation (or lack thereof) for the student's crisis.\u003c/p\u003e\u003cp\u003eAcross these solutions, certain conditions stood out. High Negative Emotions (NE) was part of nearly every configuration \u0026ndash; confirming that emotional distress is a common denominator in dropout pathways. However, as our analysis showed, NE alone was not sufficient; it was always accompanied by precipitating factors like overload or isolation. Lack of Support (LS) and Isolation (IS) each appeared in at least one major solution, indicating that either academic or social support deficits can facilitate dropout, albeit through different emotional routes (acute frustration vs. chronic apathy, respectively). Overload (OL) was critical in one of the highest coverage paths, aligning with the notion that excessive demands often directly push students away. Poor Regulation (PR), while harder to measure directly, emerged as an important internal condition explaining why some students quit even under not-so-extreme external conditions.\u003c/p\u003e\u003cp\u003eThe solutions also exhibit \u003cb\u003eequifinality\u003c/b\u003e: Solution 1 and Solution 2, for example, describe two very different profiles of at-risk students (stressed-overwhelmed vs. isolated-bored). Both profiles can lead to dropout, implying that interventions need to be multifaceted. Another observation is the idea of \u003cb\u003econjunctural causation\u003c/b\u003e \u0026ndash; it's the conjunction of factors that matters. Overload by itself was not enough to guarantee dropout if, say, support was present or the student coped well; isolation by itself might be endured if the student still found the content engaging or had strong personal motivation. It's when these factors co-occur without counterbalancing positives that dropout occurs.\u003c/p\u003e\u003cp\u003eFrom a coverage standpoint, the combination of the identified solutions explained a majority of the dropout cases in our sample (solution coverage\u0026thinsp;~\u0026thinsp;0.70, meaning 70% of dropout instances had at least one of these configurations present). A few cases remained outliers, which upon examination had very unique circumstances (for instance, one student dropped out mainly due to a sudden job offer \u0026ndash; a case of opportunity rather than negative experience). Such cases remind us that no model captures 100% of human behavior, but they were exceptions.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003e5.3 Configurations for Persistence (Contrast Analysis)\u003c/h2\u003e\u003cp\u003eAs a complementary analysis, we also briefly examined what configurations were associated with \u003cb\u003estudents persisting (not dropping out)\u003c/b\u003e despite challenges. While not the primary focus of our study, this provides a mirror image to interpret results. Persisting students tended to have at least one of the following protective patterns: \u003cb\u003estrong support\u003c/b\u003e (~\u0026thinsp;LS\u0026thinsp;=\u0026thinsp;0) \u003cem\u003eor\u003c/em\u003e \u003cb\u003elow workload\u003c/b\u003e (~\u0026thinsp;OL\u0026thinsp;=\u0026thinsp;0) \u003cem\u003eor\u003c/em\u003e \u003cb\u003eeffective coping\u003c/b\u003e (~\u0026thinsp;PR\u0026thinsp;=\u0026thinsp;0) \u0026ndash; and usually more than one of these in combination. In other words, to \u003cem\u003enot\u003c/em\u003e drop out, it helped to have multiple positives: e.g., a manageable workload and good instructor support and perhaps only mild negative emotions. This asymmetry (dropout can be caused by any one severe pathway, whereas persistence often requires all-clear on many fronts) is a sobering insight. It highlights why dropout rates are often high \u0026ndash; there are many ways to drop out (many causal recipes), but to stay consistently engaged, a student ideally needs a more complete support system and a bit of luck in avoiding major stressors.\u003c/p\u003e\u003cp\u003eTo summarize the results: \u003cb\u003eour fsQCA identified multiple sufficient pathways to intermittent dropout, each involving a mix of negative emotions and contributing factors\u003c/b\u003e. These findings validate that \u003cb\u003eno single factor\u003c/b\u003e is responsible for online dropout; instead, it is the \u003cem\u003econfluence\u003c/em\u003e of academic, social, and emotional factors that creates the conditions for a student to disengage. In the next section, we discuss these findings in light of existing theories and research, and explore what they mean for educators and institutions aiming to improve student retention in technology-mediated learning.\u003c/p\u003e\u003c/div\u003e"},{"header":"6 Discussion and Implications","content":"\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\u003ch2\u003e6.1 Discussion\u003c/h2\u003e\u003cp\u003eOur findings shed light on the \u003cb\u003emechanisms through which negative emotions drive students to intermittently drop out\u003c/b\u003e of online courses, and they reinforce the idea that dropout in technology-mediated environments is a \u003cb\u003econfigurational phenomenon\u003c/b\u003e. In this discussion, we interpret each major pathway identified, connect it to theoretical frameworks and prior research, and then delve into the broader implications for theory and practice.\u003c/p\u003e\u003cp\u003e\u003cb\u003ePathway 1: Overload\u0026thinsp;+\u0026thinsp;No Support \u0026rarr; Anxiety/Frustration \u0026rarr; Dropout.\u003c/b\u003e This pathway aligns well with \u003cb\u003einformation overload theory\u003c/b\u003e and the concept of \u003cb\u003eacademic burnout\u003c/b\u003e. When students face an excessive volume of content or tasks without adequate guidance, they can experience cognitive overwhelm that quickly turns into intense anxiety and frustration. Our results echo the findings of Chen et al. (2024) who warned that too much content and interaction can backfire, causing stress and \u003cb\u003elowering learners\u0026rsquo; willingness to continue\u003c/b\u003e. In our data, overload was not acting alone; it became truly perilous when \u003cb\u003ecombined with lack of support\u003c/b\u003e. This pairing fits with \u003cb\u003eDigital Stress\u003c/b\u003e models: stressors (like overload) coupled with lack of resources (like support) yield strain (negative emotions) and eventual withdrawal (dropout). Theoretically, this underscores the importance of \u003cb\u003eP-E fit (Person-Environment fit)\u003c/b\u003e in online learning \u0026ndash; a mismatch between demands and support triggers negative affect and maladaptive outcomes. From an \u003cb\u003eemotion regulation\u003c/b\u003e standpoint, even a student with decent coping skills can be overwhelmed if the objective demands far outstrip their capacity; no amount of deep breathing will help if you simply have 48 hours of work to do in 24 hours. Therefore, while we encourage teaching students coping techniques, \u003cb\u003ethe onus is also on course designers to manage workload and information flow\u003c/b\u003e to prevent such overload situations.\u003c/p\u003e\u003cp\u003e\u003cb\u003ePathway 2: Isolation \u0026rarr; Boredom/Disinterest \u0026rarr; Dropout.\u003c/b\u003e This configuration highlights the \u003cb\u003esocial and emotional void\u003c/b\u003e that can occur in online education. It maps onto literature about \u003cb\u003eemotional engagement\u003c/b\u003e and \u003cb\u003ebelonging\u003c/b\u003e: students who feel disconnected from peers and instructors often fail to develop investment in the course. Over time, this becomes \u003cb\u003eboredom and apathy\u003c/b\u003e, which are emotions strongly linked to disengagement. Prior research by Li and Yang (2025) found that loneliness in online learning is a significant contributor to burnout; our findings deepen that by showing loneliness (isolation) can specifically breed boredom and loss of purpose, precipitating dropout. This resonates with the \u003cb\u003eself-determination theory\u003c/b\u003e notion that relatedness is a basic human need \u0026ndash; when not met, motivation suffers. In on-campus settings, even a dull lecture can be made bearable by the presence of friends or the college environment; online, a dull course with no community feels especially pointless. Importantly, this pathway did not necessarily involve high workload or intense anxiety \u0026ndash; it was a \u003cem\u003eslow erosion\u003c/em\u003e of engagement rather than an acute crisis. The implication for theory is that \u003cb\u003edropout can stem from absence of positive emotion or stimulation\u003c/b\u003e, not just presence of negative stress. Practically, it calls for integrating \u003cb\u003ecommunity-building and interaction\u003c/b\u003e in online courses. Even simple interventions like prompt instructor feedback, synchronous meet-ups, or group projects could mitigate feelings of isolation. Research shows that fostering social presence can significantly improve engagement and reduce attrition, as \u003cem\u003e\u0026ldquo;social presence relieves burnout\u0026rdquo;\u003c/em\u003e by alleviating loneliness. Our results strongly support that: none of the well-connected students followed this boredom path.\u003c/p\u003e\u003cp\u003e\u003cb\u003ePathway 3: Poor Emotion Regulation \u0026rarr; Unchecked Negative Emotions \u0026rarr; Dropout.\u003c/b\u003e This highlights the role of the student\u0026rsquo;s \u003cb\u003einternal capacities\u003c/b\u003e. Even when external conditions were not extreme, some students fell into a dropout pattern because they couldn\u0026rsquo;t manage moderate challenges \u0026ndash; small setbacks snowballed. This finding speaks directly to \u003cb\u003eemotion regulation theory\u003c/b\u003e: those who lacked effective coping strategies (e.g., they engaged in rumination, avoidance, or denial) allowed their negative emotions to accumulate unchecked. The case of a student catastrophizing a single bad grade into \u003cem\u003e\u0026ldquo;I\u0026rsquo;m not cut out for this\u0026rdquo;\u003c/em\u003e is a prime example. Zhao et al. (\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) demonstrated that students using maladaptive strategies like expressive suppression ended up with more anxiety and poorer outcomes, which dovetails with what we saw: suppressors (or generally poor regulators) in our sample often ended up disengaging because their anxiety or frustration became overwhelming. This pathway suggests that \u003cb\u003epersonal resilience factors\u003c/b\u003e are an important part of the dropout equation. It also partially answers why some students weather the same storm that sinks others. From a theoretical viewpoint, it invites a more nuanced model of online persistence that incorporates emotional self-regulation as a moderator between challenges and outcomes.\u003c/p\u003e\u003cp\u003eInterestingly, our configurational approach suggests that improving emotion regulation alone might prevent certain dropouts, but not all \u0026ndash; because some paths (like the overload one) might overpower even good regulators. Conversely, even a perfectly designed course can lose students who are unprepared to deal with any emotional discomfort. Thus, \u003cb\u003eresponsibility is dual\u003c/b\u003e: institutions should \u003cb\u003etrain students in coping and mindset\u003c/b\u003e (for example, workshops on time management, stress reduction, fostering a growth mindset to handle failure) and simultaneously \u003cb\u003edesign courses that are supportive and humanized\u003c/b\u003e to reduce undue stress. This dual approach tackles both sides of the emotion equation. Our results empirically back the argument made by some educators that \u0026ldquo;we need to teach emotional skills, not just content\u0026rdquo; \u0026ndash; something especially pertinent in self-driven online learning.\u003c/p\u003e\u003cp\u003e\u003cb\u003ePathway 4 and others: Additional nuances.\u003c/b\u003e We noted an external-stressor path (life event\u0026thinsp;+\u0026thinsp;inflexibility) which aligns with research noting that many online learners are adults juggling multiple roles; life often intervenes. While not a primary focus, it reminds us that \u003cb\u003estructural flexibility\u003c/b\u003e (e.g., self-paced options, lenient deadlines in genuine emergencies) can catch those falling due to outside reasons. Another pattern in persistence we observed is that to avoid dropout, often multiple positive factors must coincide (the student must \u003cem\u003enot\u003c/em\u003e be overloaded, \u003cem\u003eand feel supported\u003c/em\u003e, \u003cem\u003eand\u003c/em\u003e manage emotions well, etc.). This asymmetry is worth reflecting on: it implies that \u003cb\u003epreventing dropout is harder than causing it\u003c/b\u003e, because one weak link in the chain (one risk factor) can trigger disengagement, whereas preventing it requires shoring up all links. This may explain why dropout rates remain high \u0026ndash; it's easier for something to go wrong than for everything to go right. The implication is that institutions should implement \u003cb\u003eredundant retention strategies\u003c/b\u003e: assume that just fixing one issue (say, providing tutors) may not be sufficient by itself; a combination of improvements (better course design, better support, student skills training, etc.) is needed to truly curb dropout rates.\u003c/p\u003e\u003cp\u003e\u003cb\u003eComparison with Prior Studies\u003c/b\u003e: Our results both confirm and extend prior empirical findings on online dropout. The systematic review by Rahmani et al. (\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) identified numerous factors like anxiety, isolation, and workload issues as contributors to dropout. We corroborate those factors \u003cem\u003eand show how they interlink\u003c/em\u003e. For instance, Rahmani et al. noted anxiety and concentration problems during the pandemic as major issues; our overload pathway is one concrete route to anxiety-driven dropout. Respondek et al. (\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) found anxiety predictive of dropout in traditional university; we again echo that but embed it in a causal combination. Our study also relates to research on \u003cb\u003eacademic emotions\u003c/b\u003e: Pekrun\u0026rsquo;s control-value theory suggests that emotions like boredom and anxiety arise from students\u0026rsquo; appraisals of a situation (value and control). Our isolation-boredom path suggests students saw little value (no social or intrinsic value) and possibly low control (stuck in a boring environment they couldn\u0026rsquo;t change), leading to disengagement \u0026ndash; aligning with that theory. Meanwhile, the overload-frustration path can be seen through a control-value lens as well: those students likely felt \u003cb\u003elow control\u003c/b\u003e (over the immense workload) and high value might not salvage them because they were simply incapable of controlling outcomes, resulting in anxiety (a negative activating emotion) which led to giving up.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\u003ch2\u003e6.2 Implications\u003c/h2\u003e\u003cp\u003eThe insights from this study can inform several concrete actions for educators, instructional designers, and academic policy makers:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eOptimize Workload and Content Delivery\u003c/b\u003e: Given the clear risk of information overload, online courses should be carefully moderated in terms of content density and frequency of assessments. Quality should be favored over quantity. Instructors might break content into manageable chunks, provide clear guidance on study time expectations, and use adaptive release (staging content rather than dumping everything at once). Periodic check-ins can gauge if students feel overwhelmed. As Chen et al. (2024) note, more content is not always better \u0026ndash; our data show it can be worse. Thus, content richness should be paired with equally rich support to avoid overload.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eEnhance Instructor Presence and Support\u003c/b\u003e: The lack of support factor implies that \u003cem\u003eresponsiveness\u003c/em\u003e is key. Simple practices like replying to questions within 24\u0026ndash;48 hours, holding virtual office hours, and proactively contacting inactive students can make a difference. Some institutions employ early alert systems that flag when a student hasn\u0026rsquo;t logged in or submitted work; instructors (or support staff) can then reach out in a supportive, non-punitive way (\u0026ldquo;We noticed you haven\u0026rsquo;t been active, is everything okay? How can we help?\u0026rdquo;). Establishing this net can catch students on the brink of dropout. Additionally, providing timely and constructive feedback on assignments helps reduce anxiety and uncertainty about progress.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eBuild Social Interaction and Community\u003c/b\u003e: To combat isolation and boredom, course designers should integrate discussion forums, group projects, peer feedback, or live webinars where students can see and hear each other. Even if a course is asynchronous, including an initial introduction forum, collaborative activities, or study buddy systems can foster peer connections. In our findings, those who had \u003cem\u003esomebody\u003c/em\u003e \u0026ndash; be it a friend, a mentor, or an engaged instructor \u0026ndash; were far less likely to disappear silently. Institutions might also encourage formation of student-led study groups or use technology (like social learning platforms) to create a sense of campus-like community online. These efforts are supported by evidence that \u003cb\u003esocial engagement can dramatically increase completion likelihood\u003c/b\u003e (one stat suggests students engaged in communities are \u003cem\u003e5x more likely to complete the course\u003c/em\u003e \u0026ndash; although context-specific, it illustrates the magnitude).\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eDevelop Students\u0026rsquo; Emotional and Self-Regulation Skills\u003c/b\u003e: Many students enter online learning without prior experience in that mode, and they might lack the self-regulation habits needed. Orientation modules that teach \u003cb\u003etime management, goal setting, and emotional self-awareness\u003c/b\u003e could be made standard. Furthermore, integrating brief content on stress management (e.g., the concept of growth mindset, normalizing that it's okay to struggle and seek help, tips on how to reframe challenges) could empower students. Some innovative approaches include mindfulness exercises tailored for students, or apps that check in on student well-being. Since our results show poor coping can lead to dropout, equipping students with even basic cognitive-behavioral tools (like how to challenge negative thoughts, how to break large tasks into smaller ones, etc.) is a preventative strategy. Academic advisors and counselors also play a role: reaching out to students who seem distressed and directing them to resources can prevent silent suffering that ends in withdrawal.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eFlexible Policies for Genuine Crises\u003c/b\u003e: Institutions should consider policies that allow \u003cem\u003etemporary breaks\u003c/em\u003e or \u003cem\u003eincompletes\u003c/em\u003e in courses when students face significant life disruptions. If a student knows they can take a short hiatus without penalty and resume the course, they might be less likely to fully drop out. This could be formalized as an \"intermittent study plan\" for those who need it. Our external stressor cases suggest that rigidity can turn a solvable interruption into a permanent dropout. Flexibility, coupled with support to get back on track, may convert some dropouts into persisters.\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\u003ch2\u003e6.3 Limitations\u003c/h2\u003e\u003cp\u003eIt is important to acknowledge the limitations of our study. First, the sample size and context: we had 20 students, mostly from a single institution/program. While QCA does not require large N and is suitable for small samples, the generalizability of specific configurations may be limited. Different contexts (e.g., K-12 online learning, corporate e-learning) might have other salient factors. Our study was also cross-sectional/retrospective in nature \u0026ndash; we relied on students recalling their experiences, which may introduce recall bias. Additionally, while we tried to capture the temporal aspect (what led to what), fsQCA as applied here is not a dynamic analysis; it treats the presence of conditions and outcome in a static way. Dropout is a process that unfolds, and our representation simplifies that timeline. Another limitation is measurement: calibrating qualitative data into fuzzy sets involves subjective judgment. We mitigated this with careful definitions and double coding, but there is always some imprecision. Also, some constructs like \"poor emotion regulation\" were inferred rather than directly measured with a psychometric instrument, which could be improved in future work.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e\u003ch2\u003e6.4 Conclusion\u003c/h2\u003e\u003cp\u003eOverall, our study contributes a more nuanced narrative to the story of online student dropout. It confirms that negative emotions are not just afterthoughts but are central to the dropout process, and it specifies how those emotions come about through particular combinations of stressors and lacks. It also confirms recent research that no one-size-fits-all explanation exists for dropout; instead, educators should envision multiple \u0026ldquo;persona\u0026rdquo; of at-risk students (the overwhelmed one, the isolated one, the un-coping one, etc.), each needing a tailored response. This configurational understanding encourages a shift from trying to single out \u003cem\u003ethe\u003c/em\u003e cause of dropout, to designing ecosystems that minimize multiple risks simultaneously and bolster multiple forms of support.\u003c/p\u003e\u003c/div\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003efsQCA\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eFuzzy\u0026ndash;set Qualitative Comparative Analysis\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eERT\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eEmotional Regulation Theory\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eCOT\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eCognitive Overload Theory\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eMOOCs\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eMassive Open Online Courses\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eLMS\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eLearning Management System\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eIDB\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eIntermittent Discontinuance Behavior\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv class=\"DefinitionListEntry\"\u003e\u003cdiv class=\"Term\"\u003eP\u003c/div\u003e\u003cdiv class=\"Description\"\u003e\u003cp\u003eE Fit\u0026ndash;Person\u0026ndash;Environment Fit\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003ch2\u003eEthics approval and consent to participate\u003c/h2\u003e\u003cp\u003e This study was conducted in accordance with the Declaration of Helsinki, ensuring compliance with all ethical standards pertaining to research involving human participants. Informed consent was obtained from all participants prior to their involvement in the study. Ethical approval was granted by the Ethics Committee of Tongling Polytechnic (Approval No. 5).\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003cp\u003eNot applicable.\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003ch2\u003eCompeting Interests\u003c/h2\u003e\u003cp\u003eThe authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e\u003cp\u003eThis study was also sponsored by the following:\u003c/p\u003e\u003cp\u003e1. Middle-aged and young teachers training Action: cultivating outstanding young teachers\u0026rsquo; program of Anhui Province (Grant No.: YQYB2023153).\u003c/p\u003e\u003cp\u003e2. Advanced teacher program of Tongling Polytechnic (Grant Name: Yao Yao) .\u003c/p\u003e\u003cp\u003e3. A Study on the Articulation Teaching of Specialized English for Cross-border E-commerce between Secondary Vocational and Higher Vocational Education (Grant No: tlpt2025jyzd06)\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eYao Yao: Conceptualization; Design; Drafting manuscript; Mingda Wang: Critical version of the manuscript; Supervision. Yang Zhang: Supervision.\u003c/p\u003e\u003ch2\u003eAcknowledgements\u003c/h2\u003e\u003cp\u003eThis work was supported by Tongling Polytechnic, China and Universiti Kebangsaan Malaysia, Malaysia.\u003c/p\u003e\u003ch2\u003eAvailability of data and materials\u003c/h2\u003e\u003cp\u003eThe datasets generated and analyzed during the current study are openly available in figshare at DOI:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.6084/m9.figshare.29517794\u003c/span\u003e\u003cspan address=\"10.6084/m9.figshare.29517794\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAbdelhalim SM. From Traditional Writing to Digital Multimodal Composing: Promoting High School EFL Students\u0026rsquo; Writing Self-Regulation and Self-Efficacy. Comput Assist Lang Learn. 2024;55(1):1\u0026ndash;30. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/09588221.2024.2322148\u003c/span\u003e\u003cspan address=\"10.1080/09588221.2024.2322148\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAhmed B, Yaru H, Rafiq M, Im C. Escaping the Boredom Trap: Enhancing Usability With 3D Games in Monotonous Subjects. IEEE Access. 2024;12:180361\u0026ndash;84. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/ACCESS.2024.3434673\u003c/span\u003e\u003cspan address=\"10.1109/ACCESS.2024.3434673\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003e\u0026Aacute;lvarez I, Manero B, Romero-Hernandez A, Cardenas M, Isabel, Mas\u0026oacute;. Virtual Reality Platform for Teacher Training on Classroom Climate Management: Evaluating User Acceptance. Virtual Reality. 2024;28. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s10055-024-00973-6\u003c/span\u003e\u003cspan address=\"10.1007/s10055-024-00973-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOvidiu Ioan M. To Be (Online) or Not to Be? The Antecedents of Online Study Propensity and e-Learning-Dependent Dropout Intention in Higher Education. Technol Forecast Soc Chang. 2024;207:123566. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.techfore.2024.123566\u003c/span\u003e\u003cspan address=\"10.1016/j.techfore.2024.123566\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBahtilla M. 2024. \u0026lsquo;Supervisory Feedback: Supervisors\u0026rsquo; Reasons for Not Giving Timely Feedback\u0026rsquo;. \u003cem\u003eInnovations in Education and Teaching International\u003c/em\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBalt\u0026agrave;-Salvador Ros\u0026oacute;, Olmedo-Torre N, Pe\u0026ntilde;a M, Ana-In\u0026eacute;s R-D. Academic and Emotional Effects of Online Learning during the COVID-19 Pandemic on Engineering Students. Educ Inform Technol. 2021;26(6):7407\u0026ndash;34. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s10639-021-10593-1\u003c/span\u003e\u003cspan address=\"10.1007/s10639-021-10593-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBeard KS. Flow under Threat: Stereotype Threat and Flow Theory Integration in Support of Marginalized Student Academic Engagement and Psychological Well-Being. J Educ Psychol. 2025. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1037/edu0000951\u003c/span\u003e\u003cspan address=\"10.1037/edu0000951\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBehr A, Giese M, Teguim H, Theune K. Dropping out of University: A Literature Review. Rev Educ. 2020;8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/rev3.3202\u003c/span\u003e\u003cspan address=\"10.1002/rev3.3202\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBekker CI, Rothmann S, and Magdalena M. Kloppers. The Happy Learner: Effects of Academic Boredom, Burnout, and Engagement. Front Psychol. 2023;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fpsyg.2022.974486\u003c/span\u003e\u003cspan address=\"10.3389/fpsyg.2022.974486\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBergdahl N. Engagement and Disengagement in Online Learning. Comput Educ. 2022;188:104561. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.compedu.2022.104561\u003c/span\u003e\u003cspan address=\"10.1016/j.compedu.2022.104561\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBoekaerts M. Emotions, Emotion Regulation, and Self-Regulation of Learning: Center for the Study of Learning and Instruction, Leiden University, The Netherlands, and KU Leuven\u0026rsquo;. in \u003cem\u003eHandbook of Self-Regulation of Learning and Performance\u003c/em\u003e. Routledge; 2011.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChang C-C, Szu-Ting Y. Interactive Effects of Scaffolding Digital Game-Based Learning and Cognitive Style on Adult Learners\u0026rsquo; Emotion, Cognitive Load and Learning Performance. Int J Educational Technol High Educ. 2023;20(1):16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s41239-023-00385-7\u003c/span\u003e\u003cspan address=\"10.1186/s41239-023-00385-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCloude EB, Wortha F, Dever DA. and Roger Azevedo. 2021. \u0026lsquo;Negative Emotional Dynamics Shape Cognition and Performance with MetaTutor: Toward Building Affect-Aware Systems\u0026rsquo;. Pp. 1\u0026ndash;8 in \u003cem\u003e2021 9th International Conference on Affective Computing and Intelligent Interaction (ACII)\u003c/em\u003e. New York: Ieee.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCobo-Rend\u0026oacute;n R, Hojman V, Garc\u0026iacute;a-\u0026Aacute;lvarez D, Ramon Cobo R. Academic Emotions, College Adjustment, and Dropout Intention in University Students. Front Educ. 2023;8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/feduc.2023.1303765\u003c/span\u003e\u003cspan address=\"10.3389/feduc.2023.1303765\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDelahunty J, Verenikina I. and Pauline Jones. 2014. \u0026lsquo;Socio-Emotional Connections: Identity, Belonging and Learning in Online Interactions. A Literature Review\u0026rsquo;. \u003cem\u003eTechnology, Pedagogy and Education\u003c/em\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDerakhshan A, Kruk M, Mehdizadeh M, Mirosław P. Boredom in Online Classes in the Iranian EFL Context: Sources and Solutions. System. 2021;101:102556. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.system.2021.102556\u003c/span\u003e\u003cspan address=\"10.1016/j.system.2021.102556\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEberle J, and Joyce Hobrecht. The Lonely Struggle with Autonomy: A Case Study of First-Year University Students\u0026rsquo; Experiences during Emergency Online Teaching. Comput Hum Behav. 2021;121:106804. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.chb.2021.106804\u003c/span\u003e\u003cspan address=\"10.1016/j.chb.2021.106804\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEden AL, Johnson BK, Reinecke L, Grady SM. Media for Coping During COVID-19 Social Distancing: Stress, Anxiety, and Psychological Well-Being. Front Psychol. 2020;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fpsyg.2020.577639\u003c/span\u003e\u003cspan address=\"10.3389/fpsyg.2020.577639\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEres R, Lim MH, Lanham S, Jillard C, and Glen Bates. 2021. \u0026lsquo;LonelinessEmotion Regulation: Implications of Having Social Anxiety Disorder\u0026rsquo;. Australian J Psychol.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGross JJ. Emotion Regulation: Past, Present, Future. Cognition Emot. 1999. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/026999399379186\u003c/span\u003e\u003cspan address=\"10.1080/026999399379186\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHe Y, Ding M, Liu M, Wu L, Liu Z, Guo H, Wang H. Teaching Reform and Practice Based on Four Dimensions and One Penetration for Sensing and Detection Technology. Int Rev Res Open Distrib Learn. 2024;25:66\u0026ndash;86. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.19173/irrodl.v25i4.7689\u003c/span\u003e\u003cspan address=\"10.19173/irrodl.v25i4.7689\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHeng R, Cheng Q, Pu L, Tang Q. Pre-Service EFL Teachers\u0026rsquo; Emotional Experiences, Emotion Regulation Strategies and Influencing Factors in Research: A Chinese Case Study. Heliyon. 2024;10(17). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.heliyon.2024.e36843\u003c/span\u003e\u003cspan address=\"10.1016/j.heliyon.2024.e36843\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHongxia, Li, and Juan Yang. Managing Online Learning Burnout via Investigating the Role of Loneliness during COVID-19. BMC Psychol. 2025;13(1):151. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s40359-025-02419-3\u003c/span\u003e\u003cspan address=\"10.1186/s40359-025-02419-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJimenez S, Salameh ME, Arguedas, Salvador Pons B. 2018. \u0026lsquo;Beyond the Notion of \u003cem\u003ePeriphery\u003c/em\u003e An Account of Polyfunctional Discourse Markers within the Val.Es.Co. Model of Discourse Segmentation\u0026rsquo; edited by K. Beeching, C. Ghezzi, and P. Molinelli. \u003cem\u003ePOSITIONING THE SELF AND OTHERS: LINGUISTIC PERSPECTIVES\u003c/em\u003e 292:105\u0026ndash;25.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKelly LA. 2023. \u0026lsquo;Using Engagement to Interrupt Mind-Wandering and Encourage Focus During Video-Recorded Lectures in the Online Learning Environment\u0026rsquo;. Ed.D.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKoundal N, Abdalhadi A, Al-Quraishi MS, Elamvazuthi I, Moosavi MS, Guillet C, FR\u0026eacute;d\u0026eacute;ric, Merienne, Saad NM. 2024. \u0026lsquo;Effect of Interruptions and Cognitive Demand on Mental Workload: A Critical Review\u0026rsquo;. \u003cem\u003eIEEE Access\u003c/em\u003e 12:54405\u0026ndash;25. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/ACCESS.2024.3383142\u003c/span\u003e\u003cspan address=\"10.1109/ACCESS.2024.3383142\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi D. The Shift to Online Classes during the COVID-19 Pandemic: Benefits, Challenges, and Required Improvements from the Students\u0026rsquo; Perspective. Electron J E-Learning. 2022;20(1):1\u0026ndash;18. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.34190/ejel.20.1.2106\u003c/span\u003e\u003cspan address=\"10.34190/ejel.20.1.2106\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu X, and Yijia Liu. Developing and Validating a Scale of Artificial Intelligence Anxiety Among Chinese EFL Teachers. Eur J Educ. 2025;60:e12902. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/ejed.12902\u003c/span\u003e\u003cspan address=\"10.1111/ejed.12902\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLv H, Low J, Tan Siow-kian, Tang L, Li X. Factors Affecting Medical Students\u0026rsquo; Intention to Use Rain Classroom: A Cross-Sectional Survey. BMC Med Educ. 2024;24. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12909-024-05037-6\u003c/span\u003e\u003cspan address=\"10.1186/s12909-024-05037-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMacIntyre PD, Gregersen T, Mercer S. Language Teachers\u0026rsquo; Coping Strategies during the Covid-19 Conversion to Online Teaching: Correlations with Stress, Wellbeing and Negative Emotions. System. 2020;94:1\u0026ndash;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.system.2020.102352\u003c/span\u003e\u003cspan address=\"10.1016/j.system.2020.102352\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMcRae K. Cognitive Emotion Regulation: A Review of Theory and Scientific Findings. Curr Opin Behav Sci. 2016;10:119\u0026ndash;24. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.cobeha.2016.06.004\u003c/span\u003e\u003cspan address=\"10.1016/j.cobeha.2016.06.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMiranda JO, Olga Angelinetta P, Tulabut, Reymond Neal C, Cruz. The PERMA Model of Well-Being and Student Engagement amid Covid-19: A Two-Wave Study among a Sample of Filipino University Students. Psychol Sch. 2024;61(11):35\u0026ndash;51. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/pits.23286\u003c/span\u003e\u003cspan address=\"10.1002/pits.23286\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePeng P, Chen S, Hao Y, He L, Wang Q, Zhou Y, Tang Y-Y, Yang WF, Wu Q, Tieqiao Liu. Network of Burnout, Depression, Anxiety, and Dropout Intention in Medical Undergraduates. Int J Soc Psychiatry. 2023. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1177/00207640231166629\u003c/span\u003e\u003cspan address=\"10.1177/00207640231166629\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePreston T, Carr DC, Hajcak G, Sheffler J, Natalie Sachs-Ericsson. Cognitive Reappraisal, Emotional Suppression, and Depressive and Anxiety Symptoms in Later Life: The Moderating Role of Gender. Aging \u0026amp; Mental Health; 2022.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRagin CC. Redesigning Social Inquiry: Fuzzy Sets and Beyond. University of Chicago Press; 2009.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRahmani A, Mohammad W, Groot, and Hamed Rahmani. Dropout in Online Higher Education: A Systematic Literature Review. Int J Educational Technol High Educ. 2024;21(1):19. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s41239-024-00450-9\u003c/span\u003e\u003cspan address=\"10.1186/s41239-024-00450-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRespondek L, Seufert T, Stupnisky R, Nett UE. Perceived Academic Control and Academic Emotions Predict Undergraduate University Student Success: Examining Effects on Dropout Intention and Achievement. Front Psychol. 2017;8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fpsyg.2017.00243\u003c/span\u003e\u003cspan address=\"10.3389/fpsyg.2017.00243\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRoopashree M, Praveen Kumar, Pavanalaxmi NS, Prameela, Mehnaz Fathima. \u0026lsquo;Multimedia Data Mod Education\u0026rsquo;. 2024. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/9781119786443.ch9\u003c/span\u003e\u003cspan address=\"10.1002/9781119786443.ch9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSanborn F. A Cognitive Psychology of Mass Communication. 8th ed. New York: Routledge; 2022.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShi C, Yu L, Wang N, Cheng B, and Xiongfei Cao. Effects of Social Media Overload on Academic Performance: A Stressor\u0026ndash;Strain\u0026ndash;Outcome Perspective. Asian J Communication. 2020;30:1\u0026ndash;19. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/01292986.2020.1748073\u003c/span\u003e\u003cspan address=\"10.1080/01292986.2020.1748073\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSolhi M. The Impact of EFL Learners\u0026rsquo; Negative Emotional Orientations on (Un)Willingness to Communicate in In-Person and Online L2 Learning Contexts. J Psycholinguist Res. 2024;53(2). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s10936-024-10071-y\u003c/span\u003e\u003cspan address=\"10.1007/s10936-024-10071-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTian Y, Chan TJ, Liew TW, Chen MH, Huan Na Liu. 2025. \u0026lsquo;Overwhelmed Online: Investigating Perceived Overload Effects on Social Media Cognitive Fatigue via Stressor-Strain-Outcome Model\u0026rsquo;. \u003cem\u003eLibrary Hi Tech\u003c/em\u003e ahead-of\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e-print(ahead-of-print)\u003c/span\u003e\u003cspan address=\"http://-print(ahead-of-print)\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1108/LHT-03-2024-0145\u003c/span\u003e\u003cspan address=\"10.1108/LHT-03-2024-0145\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTzafilkou K, Perifanou M, Economides AA. Negative Emotions, Cognitive Load, Acceptance, and Self-Perceived Learning Outcome in Emergency Remote Education during COVID-19. Educ Inform Technol. 2021;26(6):7497\u0026ndash;521. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s10639-021-10604-1\u003c/span\u003e\u003cspan address=\"10.1007/s10639-021-10604-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTze VMC, Lia M, Daniels, Klassen RM. Evaluating the Relationship Between Boredom and Academic Outcomes: A Meta-Analysis. Educational Psychol Rev. 2016;28(1):119\u0026ndash;44. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s10648-015-9301-y\u003c/span\u003e\u003cspan address=\"10.1007/s10648-015-9301-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWu L, and Yingling Liu. Depression, Anxiety, and Stress among Vocational College Students during the Initial Stage of Post-Epidemic Era: A Cross-Sectional Study. Medicine. 2024;103(36):1\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/MD.0000000000039519\u003c/span\u003e\u003cspan address=\"10.1097/MD.0000000000039519\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXu C, Xu Y, Xu S, Zhang Q, Liu X, Shao Y, Xu X, Peng L, Li M. Cognitive Reappraisal and the Association Between Perceived Stress and Anxiety Symptoms in COVID-19 Isolated People. Front Psychiatry. 2020;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fpsyt.2020.00858\u003c/span\u003e\u003cspan address=\"10.3389/fpsyt.2020.00858\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXue C. Mitigating EFL Students\u0026rsquo; Academic Disengagement: The Role of Teachers\u0026rsquo; Compassion and Mindfulness in China. HELIYON. 2023;9(2):e13150. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.heliyon.2023.e13150\u003c/span\u003e\u003cspan address=\"10.1016/j.heliyon.2023.e13150\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYusof R, Harith NHM, Lokman A, Batau MFA, Zain ZM, Noor Hanim Rahmat. A Study of Perception on Students\u0026rsquo; Motivation, Burnout and Reasons for Dropout. Int J Acad Res Bus Social Sci. 2023;13(7):403\u0026ndash;32. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.6007/IJARBSS/v13-i7/17187\u003c/span\u003e\u003cspan address=\"10.6007/IJARBSS/v13-i7/17187\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang, Li, and Yunjie Chen. A Blended Learning Model Supported by MOOC/SPOC, Zoom, and Canvas in a Project-Based Academic Writing Course. Research-publishing.net; 2021.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhao T, Fu Z, Lian X, Ye L, Huang W. Exploring Emotion Regulation and Perceived Control as Antecedents of Anxiety and Its Consequences During Covid-19 Full Remote Learning. Front Psychol. 2021;12. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fpsyg.2021.675910\u003c/span\u003e\u003cspan address=\"10.3389/fpsyg.2021.675910\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-psychology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"psyo","sideBox":"Learn more about [BMC Psychology](http://bmcpsychology.biomedcentral.com/)","snPcode":"","submissionUrl":"","title":"BMC Psychology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"student negative emotion, Intermittent Dropout, education technology application, mechanism, fsQCA","lastPublishedDoi":"10.21203/rs.3.rs-6971678/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6971678/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study set out to unravel the complex mechanisms by which \u003cb\u003estudent negative emotions contribute to intermittent dropout\u003c/b\u003e in technology-mediated learning environments. By using a mixed-methods approach \u0026ndash; qualitative interviews followed by fuzzy-set qualitative comparative analysis \u0026ndash; we were able to identify distinct \u003cb\u003ecausal pathways\u003c/b\u003e leading to dropout, rather than attributing it to any single factor. The findings demonstrate that \u003cb\u003edropout is a multi-faceted phenomenon\u003c/b\u003e: it can result from an overload of work and lack of support provoking anxiety, from social isolation breeding apathy, from personal lapses in coping allowing frustration to fester, or often a combination of these elements. In all cases, \u003cem\u003enegative emotions are at the core of the story\u003c/em\u003e, serving as the immediate precursors to a student's decision to disengage. This underscores a key point for both theory and practice: \u003cb\u003eemotional experiences are not ancillary to academic outcomes, but fundamentally intertwined with them\u003c/b\u003e.\u003c/p\u003e","manuscriptTitle":"Unravelling the Mechanisms of Student Negative Emotions in Technology-Mediated Intermittent Dropout: A Mixed-Methods fsQCA Study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-08 10:19:45","doi":"10.21203/rs.3.rs-6971678/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-02-13T18:59:13+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"98028834053825935614185323535109414299","date":"2025-09-04T11:39:02+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"195465057398600149046825913092610924319","date":"2025-09-04T10:29:39+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-08-28T12:42:44+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-08-13T07:09:08+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-07-22T07:19:54+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-07-13T16:44:02+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Psychology","date":"2025-07-13T16:34:55+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-psychology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"psyo","sideBox":"Learn more about [BMC Psychology](http://bmcpsychology.biomedcentral.com/)","snPcode":"","submissionUrl":"","title":"BMC Psychology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"3d122478-9541-4f4c-9b40-32f21ad77625","owner":[],"postedDate":"September 8th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2025-09-08T10:19:46+00:00","versionOfRecord":[],"versionCreatedAt":"2025-09-08 10:19:45","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6971678","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6971678","identity":"rs-6971678","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}