Integrating planetary health and environmental justice into high school construction career education: protocol for a randomized controlled trial of the Ecosystem Justice Translator

Introduction: Climate change disproportionately affects Disadvantaged Communities, yet construction workforce education rarely addresses interconnected pathways linking energy efficiency, nature exposure, and public health. Green-blue infrastructure delivers co-optimized benefits: reducing building energy consumption 15-30% while decreasing heat-related mortality by approximately 3.9% per degree Celsius of urban cooling (Gasparrini et al., 2017) - epidemiological benchmarks that inform the dose-response functions embedded in the Ecosystem Justice Translator (EJT). This protocol describes, to our knowledge, the first randomized controlled trial evaluating a curriculum intervention designed to develop planetary health competencies and environmental justice awareness among high school students pursuing construction careers. Methods and analysis: This two-arm, parallel-group randomized controlled trial targets enrollment of N=200 high school students (ages 14-18) from construction career pathway programs in the San Francisco Bay Area (over-recruitment target N=250; 25% buffer for attrition). Students are individually randomized 1:1 to intervention (Community-Centered Design curriculum integrating the Ecosystem Justice Translator) or control (traditional Virtual Design and Construction curriculum), stratified by school site using block randomization. The 6-month intervention features the Ecosystem Justice Translator (EJT) - a computational system using large language models to translate community health equity concerns into quantifiable investment priorities. The EJT's 51-theme health equity taxonomy was derived from validated public health frameworks (CDC Social Vulnerability Index, EPA EJScreen, Healthy People 2030). Primary outcome is Health-Integrated Equity Consciousness Index (HI-ECI), measured at baseline, 3, 6 (primary endpoint), and 12 months. Analysis uses intention-to-treat linear mixed-effects models with random intercepts for participants. The minimum required sample (n=26 per arm; G*Power, two-tailed alpha=0.05, 80% power, Hedges' g=0.80) is exceeded by enrolled N=200, which provides >99% power at Hedges' g=0.80 and supports multi-site confirmatory factor analysis. Ethics and dissemination: This protocol has been approved by Stanford University Institutional Review Board (IRB eProtocol #84369, approved February 13, 2026). Parental consent from a parent or guardian and written assent from each student participant are required prior to enrollment. All instruments, curriculum materials, and EJT source code will be released open-source under CC BY-NC-SA 4.0, permitting free use for educational, research, and non-profit purposes, concurrent with primary publication. Commercial licensing may be pursued separately through Stanford University Office of Technology Licensing (OTL docket S25-565). Trial registration: ClinicalTrials.gov NCT07315919. Pre-results. Protocol version 4.1, July 2026.