LUMANA Extended Data

This contains LUMANA extended data, including data collection instruments and supplementary material.The data collection instruments are provided to enhance transparency, facilitate reproducibility, and support the interpretation and potential reuse of the study methods by researchers, implementers and developers working on culturally adapted artificial intelligence (AI) systems for mental health screening in fragile settings.The repository includes the complete Kobo/XLSForm versions of the five data collection instruments used throughout the validation study. These instruments document the study workflow from participant enrolment and informed consent, assessment of cultural and linguistic appropriateness, collection of participant reflections on trust and acceptability, Hausa Idioms of distress mapping and observation of human–AI interaction within simulated clinical workflows,. Together, they capture the methodological framework used to evaluate the safety, contextual relevance, and operational feasibility of the LUMANA system.The instruments include:the participant consent and safeguarding form used to document informed consent and eligibilitythe trust and acceptability reflection instrument used to capture participants' perceptions.AI output safety assessmentthe Hausa idioms of distress mapping instrument developed to assess recognition and interpretation of culturally specific expressions of psychological distressthe human–AI workflow observation instrument used to evaluate how the AI system integrates into routine clinical decision-making processes.These materials represent the original research instruments used during data collection and are provided alongside the underlying datasets to enable readers to understand how variables were generated, how responses were collected, and how study outcomes were operationalized.The supplementary material is a complete SRQR checklist for the LUMANA manuscript.