Does participating in online communities enhance the effectiveness and experience of micro-learning? Evidence from a randomized control trial

The study addresses whether integrating online learning communities into microlearning improves learners’ knowledge acquisition and learning experience in non-formal settings. With increasing reliance on brief, bite-sized content for learning, microlearning is widely used across formal, non-formal, and informal contexts to reduce cognitive load and deliver small instructional units. While theory and prior work suggest that interactions in communities of practice could enhance learning, practical strategies for integrating online communities with microlearning are underexplored, especially in informal and non-formal environments where learners are self-directed and may lack peer support. Information overload and challenges discerning online content compound these issues. This study investigates how online learning communities, implemented via social media, affect knowledge gains and learner experience in a non-formal microlearning course on interview research methods.

Microlearning has grown with the need for flexible, lifelong learning and is prevalent in K-12, higher education, and corporate training. Prior research highlights benefits such as efficient information access, reduced cognitive load, and higher satisfaction. Beyond formal contexts, microlearning often occurs informally, where learner autonomy is high. The literature emphasizes that effective learning involves both consuming and producing content through social interaction, suggesting that peer engagement can enhance motivation and responsibility. Online learning communities, grounded in the Community of Practice theory, have been associated with improved performance and satisfaction in formal learning contexts and professional development. However, findings are mixed for community participation: some learners prefer independent study; many remain latent in online communities; diverse backgrounds and information overload can reduce communication and recognition of relevant information. Consequently, it remains unclear whether and how community participation benefits microlearning in non-formal settings, warranting empirical evaluation of its impact on knowledge acquisition and learner experience.

Design: Mixed-methods sequential explanatory design. An initial randomized controlled trial (RCT) evaluated the impact of an online learning community on microlearning outcomes, followed by semi-structured interviews to explore underlying dynamics. Participants: 100 adults (≥18 years) were recruited online; 20 dropped out (time constraints/misaligned expectations), leaving 80 completers (13 males, 67 females; mean age 24.8). Educational levels included undergraduates (16.25%) and graduate students (46.25%). For qualitative follow-up, 10 participants (5 per group) were purposively selected considering group assignment, age, test scores, and community preferences. Intervention and materials: An 8-module microlearning course on conducting qualitative interview research, comprising 20 videos (3–10 minutes each) designed under Gagné’s nine events of instruction. Videos included reflection questions (1–3 per video). Delivery was via WeChat links: group messages for the experimental group (community) and private messages for the control group. The experimental group had a WeChat group for voluntary peer discussion and regular interaction activities; the control group could message the instructor individually. The course spanned 20 days (one video per day recommended), with flexible pacing. Learners inactive for >5 days received private reminders. Procedure and data collection: Before learning, all participants completed a pre-course survey (demographics, prior knowledge) and a 20-item knowledge pre-test. Participants were randomly assigned to experimental (community) or control (no community) groups. After learning, participants completed a post-survey (learning habits, satisfaction, mental effort, community preference; experimental group additionally reported sense of community) and an identical 20-item knowledge post-test. Ten post-course semi-structured interviews were conducted via Tencent Meeting. Attrition analyses (t-tests/non-parametric) indicated no significant differences between completers and dropouts on key variables; group balance held at baseline (gender, age, pre-test, prior knowledge). Measures: - Knowledge test: 20 multiple-choice items covering interview methods; same pre/post; content validated by two experts; pilot showed expected discrimination. Reliability: pre-test KR-20=0.69 (N=100), post-test KR-20=0.73 (N=80). - Satisfaction: 9-item, 5-point bipolar adjective scale adapted from Ritzhaupt et al., 2008; translated to Chinese; α=0.85; mean score computed. - Mental effort: 9-point self-reported mental effort scale (Paas, 1992), coded 1–9 (higher = more effort). - Sense of community (experimental group only): 20-item Rovai (2002) scale with connectedness and learning subscales; 5-point Likert scored 0–4 with reverse coding per Rovai. Reliability: connectedness α=0.76; learning α=0.80; overall α=0.84. Subscale scores summed and averaged. - Preference for learning in a community: single item (yes=1/no=0) on post-survey. Data analysis: Paired-samples t-test assessed overall pre–post knowledge gains. ANCOVA tested group differences in post-test scores controlling for pre-test. One-way ANOVAs compared groups on mental effort, satisfaction, and community preference. Descriptive statistics summarized sense of community (experimental group). Interview transcripts (in Chinese, translated to English) were reviewed to extract themes on knowledge acquisition strategies and learning experiences.

- Knowledge acquisition (overall): Significant improvement from pre-test (M=12.83) to post-test (M=15.48); t(79)=9.657, p<0.001, indicating the microlearning course effectively enhanced knowledge of interview methods. - Knowledge acquisition (between groups): ANCOVA controlling pre-test scores showed no significant difference between experimental (community) and control groups in post-test performance; F(1,76)=0.257, p=0.614. - Learning experience metrics: - Mental effort: No significant group difference; experimental M=4.38 (SD=1.60) vs control M=4.55 (SD=1.69); F(1,78)=0.226, p=0.636. - Satisfaction: No significant group difference; experimental M=4.01 (SD=0.61) vs control M=4.08 (SD=0.47); F(1,78)=0.297, p=0.587. - Preference for learning in a community: Significant difference; control preferred community more (M=0.82) than experimental (M=0.57); F(1,78)=6.270, p=0.014. - Descriptive post-test means: Experimental M=15.25 (SD=3.00); Control M=15.70 (SD=2.51). - Sense of community (experimental group): Overall mean=51.95 (SD=10.09) across 20 items; connectedness mean=24.50 (SD=5.57); learning mean=27.45 (SD=6.34). - Qualitative insights: - Study behaviors were critical predictors: learners who took notes and replayed content achieved larger gains (e.g., E3: 13→20; C4: 14→20). Applying content in practice and seeking additional resources also supported gains. - Some community members experienced information overload and muted group notifications, leading to missed information and, in some cases, decreased scores. - Control group learners expressed interest in communities but were often hesitant to post due to apprehension and message volume; they valued personalized instructor messages. - Community engagement benefited some learners by exposing them to diverse perspectives and filling knowledge gaps; disengaged community participation offered little added value.

The study addressed whether online communities enhance microlearning effectiveness and experience in non-formal settings. Microlearning itself significantly improved knowledge acquisition, aligning with prior findings on its efficacy in delivering concise, focused content. However, adding an online learning community did not yield additional gains in knowledge. This may reflect the self-contained nature of the microlearning materials, the informal setting reducing commitment, and information overload in social media communities, which can disrupt attention and reduce informational uptake. Learning experience outcomes showed no group differences in mental effort or satisfaction. Microlearning’s concise format likely keeps cognitive load low regardless of community participation. Satisfaction was high in both groups, potentially influenced by the quality of materials and instructor support; personalized instructor messages to control participants may have enhanced perceived attention. Community preferences were paradoxical: learners without a community wanted one, yet those within communities sometimes muted notifications due to excessive messaging. Hesitancy to contribute publicly, large group sizes, and diverse backgrounds may suppress active participation. For some, active engagement fostered connectedness and learning, while for others, overload reduced benefits. Implications include emphasizing effective study strategies (note-taking, review/replay) within microlearning, carefully managing community size and structure to mitigate overload, curating and summarizing key discussions, and balancing community interactions with personalized instructor support. Small, background-aligned subgroups and collaborative knowledge artifacts (e.g., shared summaries) may enhance sense of community and learning gains.

This study provides empirical evidence that while microlearning effectively improves knowledge, simply adding an online learning community does not automatically enhance knowledge acquisition or the learning experience in non-formal contexts. Personalized, learner-centered microlearning design is essential, including gathering learner preferences, aligning content with their capacity for information processing, and avoiding overcrowded communities. Community managers should foster meaningful interaction opportunities and reduce excessive lurking and message overload. Designers should encourage effective study behaviors (note-taking, review) and consider mechanisms for summarizing community discussions. Future research should explore optimal integration strategies for community features that complement microlearning’s strengths and learners’ needs.

- Sample size and attrition: Only 80 of 100 recruited learners completed the study (20% attrition). Although attrition analyses suggested minimal bias, results may have limited generalizability; future work should recruit larger samples, implement retention strategies, and conduct sensitivity analyses. - Measurement of study behaviors: The post-survey did not capture specific learning behaviors (e.g., note-taking, replaying videos) that interviews identified as influential. Future designs should include these measures to quantitatively test their effects. - Short-term assessment: Knowledge was measured immediately post-course. Longitudinal assessments during and after the course are needed to evaluate retention and changes over time. - Context and platform: Findings are situated within a WeChat-based, non-formal microlearning environment; community dynamics and information overload may differ across platforms and contexts.