Suicidal Risk Survey Dataset: Bangladeshi University Students

This dataset contains 4,004 survey responses collected from bachelor's and master's degree students across higher education institutions in Bangladesh between May and August 2025. The survey instrument was developed with psychiatrist input and covers four thematic domains: demographic and socio-economic characteristics, academic and institutional background, family and parental context, and psychological well-being indicators. The target variable encodes suicidal risk derived from a four-level ordinal item on suicidal history, consolidated into a binary outcome: Risk Positive (ideation, planning, or attempt; 67.7%) and Risk Negative (no ideation; 32.3%). All variables are categorical. Responses were collected via campus-based in-person surveys and Google Forms. Participation was voluntary, responses were fully anonymised, and a crisis helpline number was displayed throughout the instrument. This dataset is released as a public research resource in conjunction with a manuscript under review. It is intended to support reproducibility and to serve as a benchmark for machine learning research on mental health risk screening in low- and middle-income country student populations.

Note: All responses were anonymised at the point of collection. No personally identifiable information is present in this dataset. Informed consent for research use and publication was obtained from all participants.