Data Scientist, Project Success and Innovation

AIR is seeking a Data Scientist to join the Project Success and Innovation (PSI) team and the Employment and Economic Opportunity Program (EEOP), providing advanced technical expertise to support innovation development, scaling efforts, and applied research and analytics across project teams. Within PSI, the Data Scientist will contribute to AIR’s Innovation Lab by co-designing, developing, and scaling solutions that strengthen and amplify effective practices across diverse technical and operational workflows.
Within EEOP, the Data Scientist will lead end-to-end quantitative workflows, including establishing and maintaining reproducible analytic environments, managing and quality assuring complex datasets, reviewing and strengthening junior staff members’ code, and conducting advanced statistical analyses. These analyses may include causal approaches, such as experimental and quasi-experimental designs, as well as non-causal approaches, including descriptive, predictive, and explanatory modeling. The Data Scientist will also help expand AIR’s capacity to responsibly apply machine learning methods in support of both research and operational objectives, while translating technical findings into clear, actionable products for clients, partners, and other stakeholders.
Candidates hired for the position may work remotely within the United States (U.S.) or from one of our U.S. office locations. This does not include U.S. territories.
About AIR:
Founded in 1946 and headquartered in Arlington, Virginia, the American Institutes for Research (AIR) is a nonpartisan, not-for-profit organization that conducts behavioral and social science research and delivers technical assistance to address some of the most pressing challenges in the United States and globally. We generate evidence and apply data-driven solutions that expand opportunities and improve lives for all.
Responsibilities
The responsibilities for the position include:
PSI (approximately 50%)
Build and maintain shared analytic tools, including code libraries, project templates, and pipelines, that research teams across AIR can adopt and reuse.
Write documentation, training materials, and onboarding guides that make modern tooling accessible to staff with varying levels of technical experience.
Support project teams adopting new tools and workflows by answering questions, troubleshooting, and relaying recurring problems back to the team.
Prototype and test new tools and platforms, including machine learning and AI applications such as large language models, and help assess their fit and responsible use at AIR.
Help maintain deployed tools by triaging issues, fixing bugs, and tracking user feedback.
Contribute to work plans, timelines, and progress tracking for PSI activities, communicating regularly with PSI leadership.
EEOP (approximately 50%)
Conduct quantitative analyses on research and evaluation projects, including analyses for causal and non-causal research questions.
Serve as lead analyst on discrete studies or analytic components, with support from senior researchers.
Manage, clean, and quality-assure complex datasets from multiple sources.
Build and document reproducible analytic workflows and quality control procedures.
Review analytic code produced by research associates and other junior staff, and support their development in reproducible analysis and documentation practices.
Contribute to study design and analytic planning discussions.
Draft analytic sections of client reports, technical memos, and presentations, and help translate findings for non-technical audiences.
Track deliverables and timelines for assigned analytic tasks.
Contribute technical content to proposals.
Qualifications:
Education, Knowledge, and Experience
Master's or doctoral degree in a quantitative field, including statistics, computational social science, education data science, statistics, economics, quantitative psychology, quantitative methods, data science, or a related discipline, recently completed or nearing completion. Coursework or applied experience in causal inference and statistical modeling.
Experience with complex survey data, sampling, and large-scale administrative, longitudinal, or linked datasets.
Experience applying quantitative methods to education, health, or social policy questions.
Experience or familiarity with machine learning and AI methods, including applications of large language models in research.
Interest in building tools that other people use, and willingness to support them in doing so.
Familiarity with data common in applied social research, such as administrative records, longitudinal student or program data, survey data, or text.
Experience with restricted-use or confidential data.
Prior work building or maintaining code used by others.
Skills
Proficiency in R and/or Python.