Decreasing Digital Distraction in College Students: Associated Online Learning Strategies Identified by Unsupervised Data Mining Approaches

Digital Distraction in Higher Education

• The unclear boundary between productive and distracting media use contributes to attention problems, while media multitasking can intensify digital distraction.Students may multitask between learning and social media for entertainment, with potential academic-performance costs.Source: Digital Distraction in Higher Education · pp. 4–5
• Identifying factors associated with distraction is insufficient; educators need targeted learning strategies that regulate attention in digital environments.The review focuses on self-regulated learning, engagement strategies, and online learning readiness, including their underexplored interconnections with distraction.Source: Digital Distraction in Higher Education · pp. 5–6
• Learner-instructor, learner-content, and peer interactions, along with active learning and relatable content, are described as ways to support attention and limit distraction.The review links peer collaboration to accountability, interactive teaching to reduced apathy, and active learning to student interest and focus.Source: Digital Distraction in Higher Education · p. 5
• Online learning readiness includes technical, social, and communication competencies that may reduce distraction and help students use digital platforms productively.ICT self-efficacy and resilience are also described as important for adapting to technology-mediated learning environments.Source: Digital Distraction in Higher Education · p. 6

Self-regulated Learning Strategies and Digital Distraction

• Self-regulated learning (SRL) involves proactive management of cognitive, motivational, and behavioral processes toward learning goals, across forethought, performance, and reflection.Strategies include goal setting, environment structuring, task and time management, help-seeking, and self-evaluation; online contexts heighten their relevance through greater autonomy and limited direct supervision.Source: Self-regulated Learning Strategies and… · p. 6
• Goal setting and planning help prioritize tasks and allocate time, environment structuring creates barriers to distraction, and self-evaluation supports reflection on off-task behavior.Examples include silencing notifications and designating distraction-free zones.Source: Self-regulated Learning Strategies and… · pp. 6–7
• Empirical investigations of which specific SRL strategies reduce digital distraction remain scarce, particularly in online learning environments.The review presents this as a gap despite theoretically plausible links between SRL strategies and distraction reduction.Source: Self-regulated Learning Strategies and… · p. 7

Engagement Strategies and Digital Distraction

• Online engagement is often operationalized through learner-instructor, learner-learner, and learner-content interactions (Moore, 1989; Martin & Bolliger, 2018); well-designed activities may mitigate disengagement and apathy (Lin & Sun, 2024), limit distraction (Lund et al., 2012), enhance cognitive engagement, and reduce off-task behaviors (Khan et al., 2017; Muir et al., 2022).Source: Engagement Strategies and Digital… · p. 7
• Although engagement strategies support motivation in online courses, whether and how they influence digital distraction remains ambiguous.The passage identifies further analysis as needed.Source: Engagement Strategies and Digital… · p. 7

Online Learning Readiness and Digital Distraction

• Online learning readiness comprises technical, social, and communication competencies, but its relationship with digital distraction has received little study.Effective academic technology use may support focus, whereas frequent personal technology use may increase distraction.Source: Online Learning Readiness and Digital… · p. 7
• Prior studies associate better online-learning preparation with stronger self-regulated learning, including time management and help-seeking, but their joint relationship with distraction remains insufficiently addressed.The section identifies an empirical gap concerning how online readiness and self-regulated-learning competencies jointly affect distraction.Source: Online Learning Readiness and Digital… · p. 8
• Participants were conveniently recruited from a southeastern U.S. public university and had to have attended at least one university synchronous online course.They completed an online survey after reflecting on one or more such courses.Source: Online Learning Readiness and Digital… · p. 8
• 530 valid responses remained; about 91% of participants were undergraduates, averaged 22 years old, and 25% were male.The participants came from diverse majors.Source: Online Learning Readiness and Digital… · pp. 8–9, Table 1 · p. 9
• t = 0.86, p = .39 indicated no significant digital-distraction difference by gender, and gender was not a meaningful antecedent of lower distraction in the association-rule results.The authors conclude that the sample’s gender imbalance is unlikely to bias the findings.Source: Online Learning Readiness and Digital… · p. 9
• The survey used validated measures of engagement, online readiness, self-regulated learning, and distraction, and confirmatory factor analysis checked their factor structures.Scales were used without major modifications, and the factor structures were estimated with ULSMV.Source: Online Learning Readiness and Digital… · p. 10

Engagement Strategies in the Online Learning Environment Questionnaire (ESOQ)

• Martin and Bolliger (2018) constructed the ESOQ to assess college students’ perceived importance of online engagement strategies across three interaction dimensions.The dimensions were Learner-to-Instructor, Learner-to-Learner, and Learner-to-Content Engagement.Source: Engagement Strategies in the Online… · p. 11
• The ESOQ uses a 5-point Likert scale from 1 = very unimportant for me to 5 = very important for me.Source: Engagement Strategies in the Online… · p. 11
• χ2/df = 3.24, RMSEA = .065, CFI = .915, TLI = .905, and SRMR = .058; ESOQ internal-consistency coefficients exceeded 0.8.CFA indicated that the questionnaire’s three-dimensional model generally had a good fit.Source: Engagement Strategies in the Online… · p. 11

Online Learning Readiness Questionnaire (OLRQ)

• Yu (2018) developed the 20-item OLRQ to measure technical competency, social competency with instructors and classmates, and communication competency.Source: Online Learning Readiness… · p. 11
• OLRQ factors assess technology experience and integration confidence, self-efficacy for online interaction with instructors and classmates, and online communication competencies.Items use a 5-point scale from 1 = strongly disagree to 5 = strongly agree.Source: Online Learning Readiness… · p. 11
• χ2/df = 2.53, RMSEA = .054, CFI = .955, TLI = .948, and SRMR = .044; OLRQ internal-consistency coefficients were higher than 0.8.CFA indicated that the four-dimension measurement model fit the OLRQ.Source: Online Learning Readiness… · p. 11

Online Self-Regulated Learning Questionnaire (SRLQ)

• The study used Barnard et al. (2009)’s 19-item, 5-point SRLQ for online or blended college learners, covering goal setting, environment structuring, task and time management, help-seeking, and self-evaluation.Source: Online Self-Regulated Learning… · p. 12
• Reliability coefficients ranged from .637 to .858, and CFA indicated fit for the five-dimension SRLQ model (χ2/df = 2.27, RMSEA = .049, CFI = .931, TLI = .917, SRMR = .064).Source: Online Self-Regulated Learning… · p. 12

Digital Distraction Questionnaire (DDQ)

• The DDQ was a 4-item, 5-point frequency scale (1 = never, 5 = very frequently; Chen et al., 2014), with reliability .814 and acceptable CFA fit (χ2/df = 3.96, RMSEA = .075, CFI = .992, TLI = .975, SRMR = .020).Source: Digital Distraction Questionnaire (DDQ) · p. 12

Data Analysis

• The two-stage analysis used association rule mining to identify strategies associated with decreased distraction, then clustered discovered rules to find broader patterns; Figure 1 depicts the process.Source: Data Analysis · p. 12, Figure 1 · p. 13

Variables Dichotomization

• For association rule mining, variables were recorded as present or absent; gender became female/male indicators, and each 5-point questionnaire item became five rating-endorsement indicators.Source: Variables Dichotomization · p. 13
• Likert responses were also encoded as rating-option and high/low importance indicators around Neutral; distraction was split into high/low by participant mean, producing 447 binary variables.Source: Variables Dichotomization · p. 13

Association Rules Mining

• ARM derives if–then rules from frequent co-occurring learning-strategy attributes, with high or low digital distraction as the consequent.The antecedent is the strategy attribute set; support determines whether an itemset is frequent.Source: Association Rules Mining · p. 14
• Association-rule strength was evaluated using support, confidence, lift, and phi; confidence measures rule reliability as the conditional probability that the consequent occurs given the antecedent.Source: Association Rules Mining · p. 14

Extracting Significant Rules

• To identify online learning strategies associated with decreased digital distraction, ARM used lift <0.95 and phi <0, extracting significant antecedent attributes; researchers first examined single-antecedent rules, then two-antecedent rules.Source: Extracting Significant Rules · p. 15
• Chi-square tests assessed whether identified antecedent attributes appeared disproportionately among students with high or low distraction.The frequency and confidence thresholds were set at 20% and 50%, respectively.Source: Extracting Significant Rules · p. 15

Clustering of Rules

• The study clustered rules using partitioning around medoids, with normalized distance defined as 1 minus the conditional probability that rules co-occurred in the same participant.This measures co-occurrence relative to how often rules occurred independently.Source: Clustering of Rules · pp. 15–16
• Because clusters group rules rather than participants, they represent hypothesized sets of low-distraction strategies across overlapping participant subgroups; cluster number, internal validity, and stability were evaluated.The Elbow Method and Silhouette scores determined the optimal cluster count, and internal validity and stability checks assessed whether clusters were distinct and meaningful.Source: Clustering of Rules · p. 16

Descriptive Statistics

• 530 college students were split by median digital-distraction scores into high-distraction (N = 314, 59.2%) and low-distraction (N = 216, 40.8%) groups.No significant demographic differences were found between the groups (p > .05).Source: Descriptive Statistics · p. 16

Associated Strategies for Digital Distraction Decrease

• Nine SRL rules linked lower distraction with goal setting, time and task management, environment structuring, choosing less-interrupted study times, and maintaining learning standards.Students reporting lower distraction also tended to report setting study-time goals, distributing study across the week, and not compromising work quality.Source: Associated Strategies for Digital… · p. 17
• Instructor reflection opportunities and interactive synchronous sessions, optional resources for deeper exploration, and confidence using digital tools were associated with lower distraction.The table also reports learner-content strategies involving realistic scenarios and learner-instructor strategies involving clear course orientation.Source: Associated Strategies for Digital… · p. 17
• Students with fewer distractions reported less frequent peer help-seeking, including consulting knowledgeable classmates or sharing problems with peers.The summary characterizes peer interaction strategies, particularly help-seeking, as inversely associated with mitigating distraction.Source: Associated Strategies for Digital… · pp. 17–18
• Gender, age, race, and majors did not appear in significant rules co-occurring with lower distraction under the support > .20 and confidence > .50 thresholds.The authors therefore report that demographics did not characterize lower-distraction behavioral profiles under their rule-extraction thresholds.Source: Associated Strategies for Digital… · p. 18
• Lower distraction co-occurred with self-regulated learning, learner-instructor and learner-content engagement, and technical confidence, while peer interaction strategies appeared inversely associated with mitigating distraction.The authors summarize these as associative findings, not evidence that the strategies caused lower distraction.Source: Associated Strategies for Digital… · p. 18, Table 2 · p. 19
• 19 significant single-attribute rules showed negative associations with digital distraction, with lift values from 0.89 to 0.95 and phi < 0.The rules indicated that specific online learning strategies were associated with lower distraction.Source: Associated Strategies for Digital… · pp. 16–17, Table 2 · p. 19

Step 2: Two Antecedent Attributes in ARM

• Among 42 identified two-attribute rules, 24 combined an SRL strategy with engagement-strategy and OLR attributes, while 14 paired two SRL strategies; none included student demographic characteristics.Seven rules were excluded due to redundancy; less frequent pairings included two engagement strategies (3), engagement strategy with OLR (1), and two OLR attributes (1).Source: Step 2: Two Antecedent Attributes in ARM · p. 23
• SRL attributes appeared in 60% of rules and engagement attributes in 26%; disagreement with peer help-seeking also recurred in multiple rules.Prominent SRL attributes were choosing a study time with few distractions, distributing study time evenly, and maintaining high learning standards; engagement attributes included learner-learner and learner-instructor strategies.Source: Step 2: Two Antecedent Attributes in ARM · p. 23

learner-content and learner-instructor engagement.

• The findings emphasized SRL strategies as the most prominent factor associated with lower distraction, with engagement strategies and technical competencies also appearing in lower-distraction profiles.The authors presented these findings as guidance for educators designing interventions to help students manage distractions.Source: Step 2: Two Antecedent Attributes in ARM · p. 23, learner-content and learner-instructor… · p. 26
• Four clusters of 42 rules associated lower distraction with combinations of environment structuring, goal setting, time management, learner-content and instructor engagement, and technical competence.Figure 2 presents the four-cluster analysis; the reported strategy combinations most frequently included environment structuring, time management, learner-content engagement, goal setting, and learner-instructor engagement.Source: learner-content and learner-instructor… · p. 23, Figure 2 · p. 24
• Cluster 1’s 12 rules paired environment structuring and goal setting with instructor engagement and technical confidence.Students described choosing low-distraction study times, maintaining high learning standards, valuing short videos for instructor presence, and communicating effectively with instructors.Source: learner-content and learner-instructor… · pp. 24–25
• Cluster 2 contained 10 rules emphasizing learner-content engagement; students reported using online resources to explore topics deeply and applying course materials to realistic scenarios, closely aligned with environment structuring.Source: learner-content and learner-instructor… · p. 25
• Cluster 3’s 13 rules paired lower peer help-seeking with environment structuring, goal setting, instructor interaction, and confidence in technical competencies.Students reported reduced reliance on peers for solving online-course problems while prioritizing instructor interactions and technical competence.Source: learner-content and learner-instructor… · p. 25
• Cluster 4’s 7 rules centered on evenly distributing study time across the week alongside environment structuring, goal setting, and learner-content and instructor engagement.The cluster characterized distraction mitigation as disciplined time management combined with structured study environments and active engagement strategies.Source: learner-content and learner-instructor… · p. 26

OLR, followed by practical implications for educators.

• Goal setting, environment structuring, and time management emerged as significant predictors of lower digital distraction in online learning.These strategies support a more structured approach to study.Source: OLR, followed by practical… · p. 26
• Students reporting lower distraction tended to report less frequent peer-focused help-seeking, a counterintuitive association.The passage reports an association, not a causal effect.Source: OLR, followed by practical… · p. 27
• The discussion suggests that online peer channels can blur academic collaboration with casual conversations that divert attention from academic objectives.Group chats and messaging tools may serve both academic discussion and off-task socializing; notifications and social-media-like interfaces may further blur boundaries.Source: OLR, followed by practical… · pp. 27–28

• Learner-learner engagement appeared less often in lower-distraction profiles than learner-content and instructor engagement, and those students placed less emphasis on peer interaction.The findings underscore the need to design peer-collaborative activities thoughtfully.Source: a course, they may turn to peer… · pp. 28–29
• Timely feedback and interactive features were significantly associated with lower distraction.Source: a course, they may turn to peer… · p. 28
• Authentic content mitigated distractions by offering deeper topic exploration and realistic scenarios for applying course material.Source: a course, they may turn to peer… · p. 28

to perceptions of inefficiency if not thoughtfully designed (Marquez, 2023).

• Confidence using computer technologies for learning often appeared among lower-distraction students, suggesting technical competence coexisted with lower distraction; institutions can train technical competency and digital self-efficacy.Source: to perceptions of inefficiency if not… · pp. 29–30

ultimately enhancing their academic success.

• At least 67% of the 42 paired rules emphasized environment structuring, goal setting, or time management as behaviors associated with lower digital distraction.Source: ultimately enhancing their academic… · p. 30
• PAM clustering yielded four clusters: cluster 4 paired time management with structured environments, while cluster 2 paired learner-content engagement with SRL strategies.Source: ultimately enhancing their academic… · p. 30
• Cluster 3 featured low peer help-seeking and greater reliance on environment structuring, goal setting, and direct instructor engagement; all clusters combined SRL and engagement behaviors.Source: ultimately enhancing their academic… · p. 30

personalized interventions to meet the needs of diverse learners in online environments.

• Educators can raise awareness of distraction’s adverse effects and incorporate explicit self-control training into courses to support self-regulation.Source: personalized interventions to meet the… · p. 31
• Instructor-led feedback and interaction opportunities tended to appear in lower-distraction profiles, supporting interactive course designs.Source: personalized interventions to meet the… · p. 31
• Peer interactions showed mixed results: poorly designed, non-goal-directed collaboration may distract, while structured group activities could maintain focus and foster collaboration.Source: personalized interventions to meet the… · p. 31

Conclusions and Future Directions

• Goal setting, environment structuring, and time management frequently appeared with lower digital distraction, whereas peer interaction strategies did not positively correlate with distraction reduction.The study examined co-occurring strategies rather than isolated ones.Source: Conclusions and Future Directions · p. 32
• The college-student sample from one public university in southeastern America may not represent the broader population, and self-reported surveys may be biased by inaccurate perceptions or recall.The paper characterizes its findings as preliminary insights from an exploratory study.Source: Conclusions and Future Directions · p. 32
• Future research could pair self-reports with screen-time tracking or behavioral observation and use longitudinal studies to examine SRL strategies’ effects on distraction over time.Source: Conclusions and Future Directions · p. 32
• Qualitative investigations could explore why peer interactions appear less effective at mitigating distractions in online learning environments.Source: Conclusions and Future Directions · p. 32

Sensitivity analysis method

• The sensitivity analysis compared 174 participants in the lowest-distraction quartile with 150 in the highest-distraction quartile.The lowest-distraction quartile served as the outcome of interest in association rule mining.Source: Appendix A · p. 50
• Because the quartile groups were smaller than the median-split groups, the minimum support threshold was reduced from 0.2 to 0.1 while other thresholds and filters stayed unchanged.Source: Appendix A · pp. 50–51

Sensitivity analysis results

• The quartile-based analysis again identified self-regulated learning, learner-content and learner-instructor engagement, and technical confidence as significant antecedents of the lowest distraction group.The authors report that these constructs were already present in the main manuscript.Source: Appendix A · p. 51
• The analysis found no qualitatively new predictors and did not make gender, age, race, or major significant rules, consistent with the primary analysis.Source: Appendix A · p. 51