Job PostingScience Lead
Sunstone Institute is seeking an experienced science lead who will lead the day-to-day work of the data scientists. You will work closely together with the Managing Editor, who leads the journalists. The place of work is Oslo, Norway.
What you'll do
Define and maintain methodological standards across data collection, bias detection, statistical modelling, and claim extraction pipelines
- Review and critique analytical work, ensuring rigour and reproducibility before findings are published or communicated externally
- Guide researchers and data scientists on appropriate methods for each problem, from study design through to interpretation
- Identify methodological risks in ongoing projects and propose mitigations
- Stay current with relevant literature and bring best practices into the team's workflows
- Contribute to how we document and communicate uncertainty in our research outputs
- Collaborate with the data infrastructure team to ensure analytical pipelines are fit for scientific purposes
- Collaborate with the Managing Editor in evaluating and selecting projects
Who we're looking for
- A strong quantitative background, PhD or equivalent experience in a relevant field (statistics, computational social science, epidemiology, political science, ecology, or similar)
- Hands-on: you read, run, and critique analysis written in Python/pandas (Jupyter/Marimo notebooks). This is a working scientific role in a small team, not a pure oversight role; you'll be in the code, not just adjacent to it
- Deep familiarity with the methodological challenges of working with observational, scraped, or otherwise imperfect real-world data
- A track record of reviewing and critiquing analytical work, you've caught real mistakes and helped people do better science
- Experience quantifying and communicating uncertainty, and adjudicating contradictory or low-confidence evidence before it becomes a public-facing claim
- Working knowledge of bias detection in datasets and model outputs
- Comfortable working across domains, collaborating with domain experts rather than being one in every field; our threat areas span environmental, democratic, and health-related systems
- Able to communicate statistical and methodological concepts clearly to non-specialists (journalists, producers, editors)
- Fluent in open-science practice: reproducible, version-controlled analysis and transparent reporting
- Intrigued by contributing to social impact and social innovation through your work
Nice to have
- Experience with data provenance, lineage, and license compliance for reused or scraped third-party data
- Experience with NLP, information extraction, or media-analysis workflows
- Familiarity with structured or semantic representations of research findings: knowledge graphs, RDF, claim-evidence modelling
- Experience working with journalists or communication officers
- Norwegian language skills (working language is English, but Norwegian is a plus, given our national focus)