Methods, Data, and Responsible AI
Match questions with appropriate methods while preserving data authority, reproducibility, equity, and human judgment.
Method follows purpose
Methods are chosen because they can produce evidence suited to the question. Descriptive, comparative, predictive, causal, qualitative, mixed-methods, implementation, design, and educational approaches answer different kinds of questions and carry different assumptions.
A methods plan records population or corpus, inclusion logic, variables or concepts, outcomes, analysis, missingness, uncertainty, sensitivity work, and review responsibilities.
Data provenance and authority
Every dataset, document set, image collection, transcript, and derived file requires a provenance record: source, owner or steward, permitted purpose, access conditions, transformations, storage location, retention, and release authority.
De-identification lowers certain risks while leaving governance responsibilities in place. Formal human-subjects research proceeds through the appropriate protocol, ethics review, consent, institutional authorization, and data-use pathway.
Responsible AI as assisted work
Artificial intelligence may support approved brainstorming, organization, coding assistance, language refinement, search development, summarization of authorized material, or pattern exploration. Human reviewers remain accountable for sources, factual accuracy, interpretation, bias, privacy, authorship, and release.
Participants record material AI assistance, verify every consequential claim against authoritative sources, inspect outputs for fabricated content and inequitable assumptions, and keep restricted information outside unapproved tools.
Applied practice
Carry the lesson into the work.
Create a one-page methods and stewardship plan naming the question, design, data source, authority, variables or concepts, analysis, AI role, verification, equity review, reproducibility record, and release decision.
Module deliverables
- Methods sketch
- Data provenance record
- AI assistance and verification plan
- Reproducibility and equity checklist
Reflection and scored assessment
Continue the pathway
Return to the complete pathway to review progress, open the next module, or download the guidebook and working packet.
Open Professional Collaboration Curriculum