IDEAS Policy Framework
Comprehensive research conduct guidelines ensuring ethical, rigorous, and transparent data science research at UM-Flint.
Responsible data science
Our policies emphasize open science, ethical conduct, data management, and compliance with regulations. We model transparency, reproducibility, and public-good orientation in everything we do.
Integrity
Highest standards of research conduct
Openness
Transparent methods and accessible results
Ethics
Responsibility at the core of every project
Seven Core Guidelines
Our comprehensive framework for responsible research conduct at UM-Flint's Data Science Institute
Research Integrity & Transparency
No fabrication, falsification, or plagiarism. All data science work must be transparently documented and open to scrutiny.
Open Science & Data Sharing
Research outputs should be made openly available to maximize impact and reproducibility. Share code via GitHub and deposit datasets in open repositories.
Reproducibility & Data Management
All research code under version control. Follow FAIR principles - making data Findable, Accessible, Interoperable, and Reusable.
Ethical & Responsible Data Use
IRB approval required for human data. Proactively consider fairness and bias in algorithms. Ethical thinking infused at every stage.
Data Privacy & Compliance
Strict compliance with HIPAA, FERPA, and all data regulations. No sensitive data on unapproved systems. Violation results in immediate action.
Inclusive Collaboration
Respect and professionalism required. Harassment strictly prohibited. Credit shared fairly. Diverse teams produce better research.
Community Engagement
Data science for public good. Serve local Flint community needs. Research should strive for positive societal impact.
Based on Best Practices
Open Access Guidelines
All scholarly work must be openly and freely accessible. Research should be legally available to everybody, free of charge.
Ethics Framework
Ethics training required for all data science students. Research on social implications of data science is mandatory.
Code of Conduct
Welcoming and supportive environment for all. No harassment or bullying. Respectful, inclusive communication required.
Reproducible Research
Version control, thorough documentation, GitHub, R Markdown. Share code and data in public repositories when possible.
"Responsibility should be the pillar of data science. Ethical thinking and social contexts must be front and center in all data science research that can impact individuals or communities."
— Ten Simple Rules for Starting a Data Science Initiative
Detailed Requirements
Data Management Plans
Required Practices:
- •Version control (Git) for all research code
- •README files and thorough documentation
- •Jupyter/R Markdown notebooks
- •Public repositories for code and data
FAIR Principles:
- FFindable - Easy to locate by others
- AAccessible - Retrievable by identifiers
- IInteroperable - Works with other data
- RReusable - Can be used in future research
Compliance & Security
⚠️ Prohibited Data on Unapproved Systems:
HIPAA, FERPA, FISMA, ITAR/EAR protected data. Violation results in immediate account closure and data removal.
IRB Approval
Required for all human subjects research
Data Use Agreements
Must follow all data sharing terms
Secure Storage
Encryption and access control required
Required Training
All researchers must complete Responsible Conduct of Research (RCR) training specific to data science, covering:
Questions about our policies?
We're here to help you understand and comply with all research conduct guidelines.
Contact us