Indigenous Wholistic Factors Project (IWFP)

Lead Author: Valentín Quiroz de la Sierra, PhD, MSW
Tribal Affiliation: Yo'eme
Academic Affiliation: Johns Hopkins Bloomberg School of Public Health, Center for Indigenous Health
Last Updated: November 2026

Overview

The Indigenous Wholistic Factors Project (IWFP) is a community-centered, data-driven initiative aimed at advancing culturally grounded understandings of suicide risk and protective factors among Native American youth in California. This repository supports the published article:

Sierra, V. Q. (2026). Predicting Suicidal Ideation Among Native American High Schoolers in California. Archives of Suicide Research, 30(1), 80–97. https://doi.org/10.1080/13811118.2025.2490154

This project uses data from the 2019–2020 California Healthy Kids Survey (CHKS) and is guided by Indigenous Wholistic Theory and the Indigenous Computational Approach. A LASSO logistic regression model was used to identify predictors of past-year suicidal ideation among Native American high school students.

Project Goals

• Advance culturally grounded knowledge of risk and protective factors to support suicide prevention among Native American youth in California
• Apply community-centered machine learning methods through the Indigenous Computational Approach to support ethical, culturally grounded public health research
• Promote open science practices that align with Indigenous Data Sovereignty principles

Repository Contents

/data/                                # Placeholder for analysis-ready data (not included)
/rawdata/                             # Placeholder for raw CHKS data (not included)
/documentation/
  ├── calschls-2019-20-crosswalk.pdf  # Survey item crosswalk
  ├── cschls_researchsum2018.pdf      # CHKS codebook and methodology
  └── appendixa.pdf                   # Construction of predictor variables

/output/
  ├── appendixb.pdf                   # School-level clustering sensitivity analysis
  └── appendixc.pdf                   # Missing data imputation sensitivity analysis

/paper/                               # Citation for the published article

/scripts/
  ├── primary_analysis.do             # LASSO model and predictors
  ├── sensitivity_clustered.do        # Clustered schools sensitivity analysis
  └── sensitivity_imputed.do          # Missing data imputation analysis

/live_vote_demo/                      # Live session notebook using synthetic data

README.md                             # Main project documentation

Live Vote Demonstration

live_vote_demo/ holds a Python notebook used in a live session: a room votes on survey factors, and their top-voted items go into a LASSO model. It runs on synthetic data, not CHKS responses, and its results are not findings.

Software Requirements

Analyses require Stata/SE 17.0 or higher and the following user-written packages:

• lasso2 – Penalized regression
• cvauroc – AUC estimation
• calibrationbelt – Model calibration assessment
• moremata – Matrix and math utilities

Install them in Stata using:

ssc install lasso2
ssc install cvauroc
ssc install calibrationbelt
ssc install moremata

Data Availability

The California Healthy Kids Survey (CHKS) data are not publicly shared in this repository due to privacy protections for youth and school districts. Researchers may request access from WestEd or local education authorities subject to IRB approval.

OSF Link

All study materials — including supplemental documentation, appendices, and pre-registration — are archived at the Open Science Framework (OSF): 🔗 https://osf.io/4dpwt/

Contributing

This repository is intended to support transparent, ethical, and collaborative computational research in Indigenous health. If you'd like to adapt or contribute to this project:

1. Fork the repository and submit a pull request
2. Include a description of your proposed changes
3. Ensure your contributions respect Indigenous research ethics and Indigenous Data Sovereignty principles

Citation

If you use or adapt any portion of this repository, please cite:

Sierra, V. Q. (2026). Predicting Suicidal Ideation Among Native American High Schoolers in California. Archives of Suicide Research, 30(1), 80–97. https://doi.org/10.1080/13811118.2025.2490154

The Indigenous Computational Approach used in this project is described in:

de la Sierra, V. Q. (2026). A Guide to Constructing Indigenous Statistical Spaces for Prevention Science Research. Prevention Science, 27(7), 1239–1252. https://doi.org/10.1007/s11121-026-01911-5

Contact

For questions, collaboration opportunities, or data requests, contact:
Valentín Quiroz de la Sierra, PhD, MSW (Yo'eme)
Postdoctoral Fellow
Johns Hopkins Bloomberg School of Public Health
Center for Indigenous Health
vsierra4@jh.edu