Biostatistician
# Biostatistician **ISF** · Remote · `Remote` 🕒 **Статус:** *Опубликовано: 5 дней назад* · *Источник: Indeed* --- ### About the Role INDEPENDENT CONTRACTOR (1099) POSITION Consulting Biostatistician — Statistical Methods Review & Publication Support Public Health Analytics | Applied Government Research Engagement Type Independent Contractor — 1099 Reports To Principal, Management Consulting Core Function Independent methodological review of a Bayesian hierarchical / machine-learning predictive modeling pipeline; potential contribution to peer-reviewed publication of the work Education Doctorate (Ph.D. or equivalent) in biostatistics, statistics, quantitative epidemiology, or a closely related field with demonstrated Bayesian modeling experience Hours Approximately 10–20 hours per month on a defined-deliverable basis, with the option to scale hours depending on project need Availability Occasional 1-hour meetings between 9 AM – 5 PM ET (onboarding, periodic client meetings); internal deadlines agreed with the Principal at tasking Compensation Hourly rate commensurate with senior doctoral-level consulting experience; billed monthly on hours worked Tools Proficient in R and at least one Bayesian modeling framework (e.g., Stan, brms, INLA, NIMBLE, JAGS). Our code and data are hosted in Azure / Azure Databricks; we will onboard you into that environment and prior exposure is helpful but not required. Use of Claude Code for code review is encouraged. Data Access Must currently hold, or be able to obtain, CITI Human Subjects Research certification before restricted data access is granted Project Overview This contractor would support an applied public health research project producing a state-level decision-support system for a government agency client. The system is built on a two-stage predictive modeling pipeline: a Bayesian hierarchical abundance model that estimates the latent at-risk population from six surveillance outcomes across thousands of census tracts statewide, followed by a gradient-boosted machine learning layer with SHAP-based feature importance. Model outputs feed a live Azure-hosted interactive dashboard used by agency stakeholders for planning and resource allocation. A doctoral-level lead data scientist (Ph.D., Public Health) designs and manages the full analytical pipeline. A second data scientist supports data structuring and pipeline development. This Consulting Biostatistician would serve as an independent methodological reviewer and collaborator: evaluating the statistical specification and its assumptions, providing structured written feedback before deliverables reach the client, and contributing to the preparation of peer-reviewed publications describing the methods and findings. Responsibilities Statistical Methods Review Evaluate the Bayesian hierarchical model specification for internal consistency, identifiability, and appropriateness of assumptions given census-tract-level surveillance data. Assess prior specification, spatial structure (e.g., CAR/ICAR), convergence diagnostics, and posterior predictive checks; recommend sensitivity analyses where warranted. Review the machine-learning layer (gradient-boosted models, SHAP feature importance) for appropriate validation design, leakage risk, and defensible interpretation of feature importance. Advise on small-area estimation challenges, suppression and interval censoring, and ecological inference limitations, and on how these should be characterized in deliverables. Flag inconsistencies between documented methods and implemented code; read and run R / SQL scripts as needed to confirm that the documented approach is what was implemented. Technical Documentation Review Review methods and results sections of the technical deliverable as they are updated, confirming that statistical summaries and methodological descriptions accurately and clearly represent the analytical work. Identify places where results are mischaracterized, over-interpreted, or ambiguously described, and provide written review comments for the lead data scientist to address. Advising the Principal Provide an hour estimate for each task before proceeding with work. Provide brief written status updates on work in progress. Provide objective technical feedback on the model and its outputs, and advise on future methodological scoping with the client as requested. Candidate Profile Strong candidates will bring a doctoral credential in biostatistics, statistics, or quantitative epidemiology alongside meaningful hands-on experience with Bayesian hierarchical modeling and public health surveillance data. The role requires someone who can evaluate statistical assumptions with authority, give constructive, well-reasoned written feedback to a doctoral-level collaborator, and communicate methods clearly to both technical and non-technical audiences. Experience with government-contracted or externally accountable research is essential; a publication record in applied biostatistics or epidemiology is strongly preferred. Biostatistics & Bayesian Modeling Ph.D. (or equivalent) in biostatistics, statistics, quantitative epidemiology, or a closely related field. Deep, hands-on experience with Bayesian hierarchical modeling — prior specification, posterior inference, convergence assessment, posterior predictive checking — applied in real analytical contexts. Experience with spatial and small-area estimation methods (CAR/ICAR, BYM-type models) relevant to census-tract-level public health prediction. Strong grounding in health surveillance methodology, including surveillance-based models for disease burden, health outcomes, or population risk estimation. Understanding of suppression, interval censoring, and ecological inference limitations in small-area public health data. Experience with gradient-boosted models and SHAP-based feature importance is a meaningful plus. Domain Experience Experience with public health datasets — administrative health records, surveillance systems, vital statistics, or social determinants of health data — strongly preferred. Familiarity with behavioral health, substance use, or opioid-related surveillance is a plus. Experience delivering quantitative work in externally accountable contexts: consulting, applied research, or government advisory work. Peer-reviewed publications in applied statistics, epidemiology, or public health strongly preferred. Coding & Reproducibility Preferred candidates have proficiency in R, including packages for data manipulation, spatial analysis, and Bayesian modeling (e.g., rstan, brms, INLA, nimble, sf, spdep). Preferred candidates have an ability to read, run, and evaluate someone else's R / SQL code within a documented, multi-source analytical pipeline. Familiarity with version control (Git or equivalent) and reproducible, well-commented analytical code. Prior exposure to Azure, Databricks, or R Shiny is helpful but not required; onboarding to the project environment will be provided. Project Estimation & Time Management Ability to assess a defined scope of work and offer a reasonable hour estimate before beginning. Comfort surfacing scope questions and clarifying tasks early. Experience tracking and reporting hours on consulting or contract work. Client Interface & Communication Reports directly to the Principal. Ongoing technical collaboration with the lead data scientist is expected; tasking, deadlines, and client engagement are directed by the Principal. Client interaction may be requested, including deliverable walkthroughs and discussion of the model framework, inputs, and outputs. The successful candidate will represent the organization with professionalism and describe technical concepts with clarity and accuracy. Responsive and proactive in communicating blockers or schedule constraints; comfortable working in a government-contracted environment with external deadlines. Engagement Logistics IRB Compliance & CITI Training Access to project data requires current CITI Program certification in Human Subjects Research. Candidates without current certification must complete CITI training (self-paced, citiprogram.org) before data access is granted. Additional data use agreement or IRB protocol requirements will be communicated at onboarding. Independent Contractor Status This is a 1099 independent contractor engagement. The contractor is responsible for their own taxes and benefits, as well as resources required to complete work assignments, including access to a laptop, secure internet connection, and appropriate open-source statistical software. No employment relationship is created or implied. Confidentiality & Data Use The contractor will have access to sensitive public health surveillance data governed by applicable data use agreements and confidentiality obligations. Doctorate (Ph.D. or equivalent) in biostatistics, statistics, quantitative epidemiology, or a closely related field with demonstrated Bayesian modeling experience Hours Approximately 10–20 hours per month on a defined-deliverable basis, with the option to scale hours depending on project need Availability Occasional 1-hour meetings between 9 AM – 5 PM ET (onboarding, periodic client meetings); internal deadlines agreed with the Principal at tasking
Наблюдалась 2026-09-21, впервые 2026-09-16, источник — Indeed.