The study, conducted by Ziqi Jia of the School of Marxism Studies at Nanjing University of Science and Technology, surveyed undergraduates at three institutions in Xuzhou with different admission thresholds: China University of Mining and Technology, a national “Double First-Class” research university; Jiangsu Normal University, a provincial normal university; and Xuzhou University of Technology, a municipal institute. Participants were recruited through snowball convenience sampling via WeChat and QQ group chats, completed an anonymous online questionnaire on the Wenjuanxing platform, and had to be currently enrolled undergraduates aged 18 or older who were active Douyin users. Of 279 responses, eleven students who did not use the platform were excluded, leaving 268 valid cases.

The headline finding concerns prevalence. Some 67.17 percent of participants reported using Douyin for academic purposes, while 32.83 percent said they did not. Yet the content that drew them to the app was overwhelmingly recreational: humor ranked first as a preferred genre at 55.69 percent, followed closely by lifestyle documentation and vlogging at 55.49 percent. Skill-acquisition tutorials came third at 45.66 percent, making educational content a significant but secondary draw. In other words, students frequently stumble into learning on the platform, but entertainment remains the primary engine of engagement.

When asked directly about academic consequences, the responses defied simple categorization. A majority, 72.76 percent, felt Douyin had some influence on their studies, but only 18.66 percent reported predominantly positive effects, and most of those described the benefits as moderate rather than transformative. Negative effects were reported by 14.93 percent. The largest single group, 39.17 percent, gave ambivalent evaluations, describing both gains and costs. This tripartite distribution, with ambivalence outweighing either pure endorsement or pure condemnation, is central to the study’s framing of the platform as a double-edged sword whose net effect depends on individual usage patterns, the content consumed, and the user’s capacity for self-regulation.

The institutional breakdown revealed pronounced descriptive differences. At Xuzhou University of Technology, 54 percent of respondents endorsed beneficial academic effects, far exceeding the 28 percent at the more selective China University of Mining and Technology and 18 percent at Jiangsu Normal University. Negative appraisals showed the reverse concentration: 45 percent at Jiangsu Normal University, 32.5 percent at the research university, and 22.5 percent at the technology institute. The author notes that these are descriptive patterns in the sample, not statistically verified population differences, and offers self-regulatory capacity, academic socialization, and academic pressure as tentative hypotheses for future work rather than tested explanations.

Why did students credit the platform with benefits? The most commonly cited positive factors were instrumental: 56.65 percent pointed to domain-specific knowledge gains, and 45.09 percent highlighted timely access to information such as graduate entrance examination materials and civil service test updates. Other facilitators included connection with peer instructors, motivational boosts, and help modulating work and rest. On the negative side, the dominant self-reported problems were temporal displacement, meaning compulsive consumption crowding out study hours, and cognitive interference, meaning attentional fragmentation during learning episodes. A minority also cited the pull of commercial livestreaming as a competing demand on scholarly time.

Behavioral indicators added an important layer to these perceptions. Daily consumption was mostly moderate: 43.28 percent reported under one hour of use, 29.48 percent between one and two hours, 15.67 percent between two and three hours, and 11.57 percent more than three hours. The author cautions that these self-reported durations describe consumption time only and should not be read as diagnoses of addiction or problematic use. More revealing is the gap between perceived value and actual learning-oriented use. Although 45.66 percent of participants cited skill acquisition as a motivation, only 9.32 percent reported frequent educational engagement with the platform, 58.96 percent used it as a learning tool only occasionally, and 31.72 percent never did. Meanwhile, 47 percent acknowledged the power of learning videos, 36.6 percent saw no clear benefit, and 16.4 percent dismissed their pedagogical value altogether.

The most consequential numbers concern automaticity. When asked about non-deliberate use during study episodes, 33 percent reported rare spontaneous use and 46.1 percent sporadic use, but 20.9 percent admitted habitual, unconscious checking, a figure that exceeds the share reporting more than three hours of daily consumption. This separation between intentional pedagogical use and reflexive consumption suggests that the platform’s architecture, with its infinite scrolling, variable reward schedules, and algorithmically personalized content streams, may cultivate habitual checking behaviors that operate independently of conscious learning goals. Consistent with this, 87.3 percent of respondents said Douyin rarely or only occasionally interfered with their schedules, while 12.7 percent reported regular disruption, indicating that serious interference clusters in a subset of high-risk usage patterns rather than spreading uniformly across the sample.

Among participants who reported negative academic impacts, continuance intention split evenly between those who intended to keep using the platform and those who intended to stop, a result suggesting that perceived harm alone does not predict behavioral termination. The author speculates that subjective preference, optimism bias, or limited alternative coping strategies may explain this persistence, while emphasizing that the survey’s descriptive design cannot test such mechanisms. Indeed, the study is explicit about its interpretive limits: Social Cognitive Theory serves as the main framework, with the Technology Acceptance Model, Cognitive Load Theory, and Distraction Conflict Theory used only as interpretive lenses, since no validated multi-item scales for their core constructs were included in the questionnaire.

The limitations are substantial and clearly stated. Snowball convenience sampling across three Xuzhou institutions restricts external validity and may introduce bias; all data are self-reported, inviting both self-report and common-method bias; the cross-sectional design cannot establish causation; no objective academic records such as grade point average were collected; and formal reliability testing was not possible because most constructs were measured with single items rather than validated scales. Within those constraints, the study offers practical leads rather than proven interventions: digital literacy training that helps students distinguish quality educational content from fragmented entertainment, differentiated support tailored to institutional tiers, and instructor curation of high-quality short videos as supplementary out-of-class materials. The author stresses that none of these proposals constitute ready-to-deploy campus policy, and calls for longitudinal and intervention-focused research to test whether such practices measurably improve students’ learning experiences and self-regulatory behaviors on short-form video platforms.

Subject of Research: Self-reported academic consequences and self-regulatory behaviors of Douyin use among Chinese undergraduates

Article Title: The double-edged sword of short-form video platform use: Chinese undergraduates’ perceived academic consequences and self-regulatory behaviours on Douyin

Article References: Jia, Z. (2026). The double-edged sword of short-form video platform use: Chinese undergraduates’ perceived academic consequences and self-regulatory behaviours on Douyin. Frontiers in Psychology, 17, Article 1863958. https://doi.org/10.3389/fpsyg.2026.1863958

Image Credits: AI Generated

DOI: 10.3389/fpsyg.2026.1863958

Keywords: Douyin, TikTok, short-form video, Chinese undergraduates, perceived academic impact, self-regulation, microlearning, cross-sectional survey, digital literacy, educational psychology, higher education, attention