Developing Microlearning and Counseling Chatbots for Preventing Online Doxing and Shaming among University Students: Feasibility and Usability Study

Andi Nurjanna
Aan Saputra
Dewi Marlin
Ayu Qanita
A. Ratna
Indah Puteri

Information Technology Education Journal
Vol. 3, No. 2, Mei (2024)

The rapid proliferation of online harassment, particularly doxing and shaming, poses significant psychological and social risks for university students. Despite growing awareness campaigns, preventive interventions remain limited. This study aims to develop and evaluate a combined microlearning and chatbot counseling intervention to enhance students’ digital safety knowledge, preventive behavioral intentions, and engagement in higher education contexts. The key argument is that interactive, concise digital learning combined with AI-driven counseling can effectively reduce risks of online harassment. Design/methods/approach – A feasibility and usability study was conducted with 60 undergraduate students. Participants completed pre- and post-intervention assessments of knowledge and behavioral intentions, alongside a System Usability Scale (SUS) survey for chatbot evaluation. The intervention comprised bite-sized microlearning modules addressing online safety, reinforced with scenario-based interactions through a chatbot counseling system. Data were analyzed using paired t-tests, descriptive statistics, and correlation analyses to assess knowledge gains, behavioral intention changes, and engagement metrics. The intervention significantly improved participants’ knowledge scores (pretest mean = 54.2, posttest mean = 81.5, p < 0.001, Cohen’s d = 2.28) and preventive behavioral intentions (all dimensions p < 0.001). SUS scores averaged 82.3, indicating high usability, while engagement metrics revealed near-complete module completion (94%) and frequent chatbot interactions (mean 12.4 interactions per participant). Knowledge gains were positively correlated with module engagement (r = 0.61, p < 0.001). The study demonstrates the feasibility and efficacy of integrated microlearning-chatbot interventions, offering scalable preventive education strategies. Limitations include single-institution sampling, short intervention duration, and reliance on self-reported measures, which constrain generalizability. This study provides empirical evidence for combining microlearning and AI counseling in digital safety education, with potential for adaptation to other forms of online harassment or digital literacy initiatives. Future research should explore longitudinal impacts and multi-institution trials.