BCIwiki (bciwiki.com) – What decides whether Chinese college students intend to use a brain-computer interface is whether their teachers back it, not how useful they think the technology is, according to a survey of 800 students at 10 universities across eastern, central, and western China, published in Frontiers in Human Neuroscience on August 21, 2026.
The study extended the Unified Theory of Acceptance and Use of Technology, adding personal innovativeness, neuroethical concern and teacher support to the original performance expectancy, effort expectancy, social influence and facilitating conditions, all measured on 5-point scales and analysed with partial least squares structural equation modelling. Teacher support carried a path coefficient of 0.337, the only predictor reaching a moderate effect size (f-squared 0.19), followed by personal innovativeness at 0.219, effort expectancy at 0.156, and social influence at 0.121, which was only marginally significant. Performance expectancy, the centrepiece of the classical model, came in at 0.059 and did not reach significance; facilitating conditions were likewise non-significant. The model explained 62.6% of the variance in behavioural intention. Because few students have ever used a BCI directly, the authors argue, they fall back on vicarious experience, which puts teachers in the role of the nearest reference point.
The survey and the interviews diverge on one point worth dwelling on. Neuroethical concern carried a negative coefficient of 0.044 in the structural model and was not significant, yet 28 of the 40 students who sat for semi-structured interviews named privacy anxiety as their leading worry, the most frequently raised of the four barriers. The authors resolve the tension by pointing to the zero-order correlation between concern and intention, which was positive at 0.24: before anyone has actually used the technology, worry appears to track engagement rather than refusal. The questionnaire captured abstract unease, while the interviews drew out concrete demands about governance.
Beyond privacy anxiety, the interviews surfaced insufficient institutional readiness, raised by 25 students, technological unfamiliarity, raised by 22, and the absence of ethical review, raised by 19. Specific worries covered the collection, storage and secondary use of EEG data, device reliability and misclassification, the lack of campus infrastructure and technical support, and existing ethics committees that are not equipped to review neurotechnology. The recommendations that follow put teacher training and buy-in first, alongside dedicated neuroethics review, explicit data governance with withdrawal rights, and infrastructure built before deployment.
The authors set out the limits of the work. The sample skews: 82.6% of respondents came from eastern China, 58.3% were first-year students, and only 2 were in their fourth year. The cross-sectional design rules out causal inference, and the neuroethical concern and teacher support scales were adapted rather than formally validated.