Revolutionizing AI in Prevention Science
Constructing Indigenous Statistical Spaces for Culturally Grounded Research
This paper introduces the Indigenous Computational Approach (ICA), a structured protocol integrating AI-powered methods with Indigenous Research Methodologies. It addresses epistemological, technical, and ethical challenges in applying AI to Indigenous contexts, ensuring research supports Indigenous Data Sovereignty and community-defined well-being.
Key Outcomes & Strategic Impact
The Indigenous Computational Approach (ICA) provides a framework to enhance the relevance and effectiveness of AI in prevention science research, particularly within Indigenous communities.
0.0
Model Discrimination (AUC)
0.0
Model Calibration (Brier Score)
0
Features Retained by Model
0
Indigenous Suicide Rate vs. National Average
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
Researcher Standpoint
Indigenous Theoretical Framework
AI Data Analysis Technique
Dissemination & Indigenous Governance
Centering Positionality and Accountability
Researcher Standpoint in the ICA involves explicitly naming the researcher's social position, epistemological commitments, and relationships to Indigenous communities. It ensures relational accountability, defines research questions collaboratively, and documents Indigenous Data Sovereignty processes for data access and governance. This initial step frames the computational workflow to align with Indigenous worldviews, preventing biomedical assumptions from overshadowing community-defined priorities. The IWFP case study exemplifies this by grounding the research in the author's Yo'eme worldview, emphasizing 'en tui hiapsimake' ('with good heart') and partnering with Native youth ambassadors to define research questions and govern data use.
Guiding Principles for Holistic Health
The Indigenous Theoretical Framework component integrates Indigenous Knowledge Systems to conceptualize health and well-being within research projects. It ensures that prevention research is not theory-neutral but actively guided by Indigenous perspectives, emphasizing interconnectedness of physical, mental, spiritual, and environmental health. This framework dictates what constitutes meaningful constructs, how risk and protection are understood, and how statistical associations are interpreted. In the IWFP, the Indigenous Wholistic Theory guided feature selection across five domains (individual, spiritual-historical, emotional-social, mental-political, physical-economic) to capture a comprehensive understanding of well-being among Indigenous youth, moving beyond purely psycho-centric views.
Culturally Responsive Computational Methods
AI Data Analysis Technique encompasses the analytic decisions within the computational workflow—from data preprocessing to model interpretation—grounding them in Researcher Standpoint and Indigenous Theoretical Frameworks. It challenges the notion of AI models as neutral tools, highlighting that these decisions determine which forms of knowledge become visible. This component ensures transparent data preprocessing, justification of model choices based on Indigenous frameworks, co-design of evaluation procedures with governance partners, and final interpretation aligned with the Indigenous Theoretical Framework. The IWFP used lasso logistic regression, selected and approved by Native Youth Ambassadors, to identify a parsimonious, interpretable subset of predictors for suicidal ideation among Indigenous youth, ensuring alignment with the Indigenous Wholistic Theory.
Upholding Data Sovereignty and Community Benefit
Dissemination and Indigenous Governance are integral to the ICA, ensuring research outputs are governed and shared in ways that uphold Indigenous Data Sovereignty and benefit communities. This component operationalizes the CARE Principles (Collective Benefit, Authority to Control, Responsibility, Ethics) by positioning Indigenous communities as active participants in how findings are applied and disseminated. It includes co-defining acceptable use of findings, returning results in community-defined formats (e.g., town halls), establishing ownership and benefit-sharing agreements, and setting community-defined next steps. The IWFP exemplified this by holding town halls, developing a culture-based risk assessment tool, and making code publicly available only after community discussions and approval.
Enterprise Process Flow: Indigenous Computational Approach (ICA)
Locate Standpoint
→
Map Relational Accountability Processes
→
Finalize Research Questions
→
Document Data Access & Indigenous Governance Processes
→
Select Indigenous Theoretical Framework
→
Define Core Constructs
→
Map Constructs to Data
→
Set Interpretation Protocol
→
Preprocess Data with Transparency
→
Select AI Model(s) Using Components 1+2
→
Co-Design Evaluation Procedures
→
Conduct Final Interpretation
→
Co-Define Acceptable Use of Findings
→
Return Results in Community-Defined Formats
→
Establish Ownership & Benefit-Sharing
→
Set Community-Defined Next Steps
Case Study: Indigenous Wholistic Factors Project (IWFP)
The IWFP applied the ICA to predict suicidal ideation among Indigenous high school students in California (n=2609). Grounded in the 'en tui hiapsimake' (with good heart) Researcher Standpoint and the Indigenous Wholistic Theory, the project utilized lasso logistic regression for feature selection. Ten out of 17 candidate features were retained. The model demonstrated strong discrimination (AUC=0.87) and acceptable calibration (Brier score=0.10). Key findings included depressive symptoms, school-based victimization, substance use, and sexual/gender minority status as predictors. The project emphasized community governance, returning results through town halls, and co-defining next steps, such as developing a culture-based risk assessment tool for the Sacramento Native American Health Center. This case study illustrates how ICA restructures AI model design, validation, and deployment to align with Indigenous determinants of health and data sovereignty.
| Feature | Conventional AI Approach | Indigenous Computational Approach (ICA) |
|---|---|---|
| Epistemological Alignment |
|
|
| Data & Feature Selection |
|
|
| Governance & Ethics |
|
|
| Model Interpretation |
|
|
| Dissemination of Findings |
|
|
2x
Indigenous Suicide Rate Compared to National Average
Projected Impact Calculator
Estimate the potential gains from integrating the Indigenous Computational Approach into your research workflows.
Annual Research Savings
$0
Hours Reclaimed for Community Engagement
0
Your Implementation Roadmap
A typical journey to integrate Indigenous Computational Approaches into your prevention science research.
Phase 1: Community Engagement & Standpoint Definition
Establish governance partners, define research questions collaboratively, and document Indigenous Data Sovereignty processes for data access and use.
Phase 2: Framework Integration & Data Mapping
Select the Indigenous Theoretical Framework, define core constructs relevant to well-being, map these constructs to available data, and identify structural data gaps.
Phase 3: AI Model Co-Design & Validation
Preprocess data transparently, select appropriate AI models with community input, and co-design evaluation procedures to ensure cultural relevance and ethical alignment.
Phase 4: Interpretation & Indigenous Governance
Conduct final interpretation of model outputs with governance partners, co-define acceptable use of findings, establish ownership and benefit-sharing, and plan community-defined next steps.
Ready to Transform Your Research?
Book a consultation with our experts to explore how the Indigenous Computational Approach can elevate your prevention science initiatives.
Ready to Get Started?
Book Your Free Consultation.
Let's Discuss Your AI Strategy!
Lets Discuss Your Needs
Select a Date
Sun
Mon
Tue
Wed
Thu
Fri
Sat