The Griffith Lab at the University of Chicago is seeking a Data Scientist/Statistician I to support research in women's brain health and Alzheimer's disease prevention. The role will involve collecting, organizing, and analyzing data from various internal and external sources to identify risk and resilience factors for Alzheimer's, with a focus on sex-specific (female) risk. The primary challenge is to integrate and harmonize large-scale, complex datasets from multiple U.S. and international sources to create analysis-ready datasets for predictive modeling and research publications.
Requirements
- Foundational knowledge and hands-on practice in core statistical methods – descriptive inference, probability, linear/logistic regression – with implementation in R/Python and clear interpretation.
- Hands-on experience harmonizing cognitive assessment data and applying measurement invariance/IRT/score linking.
- Practical knowledge of missing data (MICE, weighting).
- Experience publishing harmonized datasets and reproducible reports (R Markdown/Quarto/Jupyter).
- Foundational knowledge and hands-on practice in survival analysis, mixed-effects models and longitudinal modeling.
- Experience with health data standards (ICD, SNOMED, LOINC, HL7 FHIR or OMOP) and unit/scale conversions (UCUM).
- Programming and Coding experience.
Responsibilities
- Implement the research analyses, executing large-scale data harmonization, statistical analysis, and modeling – including predictive models for Alzheimer’s disease with a particular focus on sex-specific (female) risk.
- Acquiring, cleaning, and organizing datasets; mapping and assessing available cohorts and sources by profiling variable coverage, coding systems, and cognitive instruments; prototyping an end-to-end harmonization on subsets of several cohorts with documented variable/value maps and quality-control checks; validating measurement invariance and IRT linking on one to two cognitive scales to produce crosswalks and uncertainty summaries; and delivering an analysis-ready, versioned dataset accompanied by a data dictionary and a concise user guide.
- Unify different types of data including cognitive instruments (e.g., MMSE, MoCA, Trails, Digit Symbol, HVLT, etc.): perform measurement invariance testing; build IRT/linking models and score crosswalks; document comparability limits.
- Correct site/batch effects and temporal drift using mixed-effects models, empirical Bayes approaches, and sensitivity analyses.
- Handle missing with principled methods (e.g., MICE, IPW); quantify robustness.
- Maintains and analyzes statistical models using best practices in machine learning, statistical inference, and reproducible research workflows.
- Builds and analyzes statistical models and reproducible data processing pipelines using knowledge of best practices in machine learning and statistical inference.
Other
- Minimum requirements include a college or university degree in related field.
- Minimum requirements include knowledge and skills developed through 2-5 years of work experience in a related job discipline.
- Graduate college or university degree.
- Masters degree in Biostatistics, Statistics, Epidemiology, Psychometrics, Data Science, or related field.
- Excellent written and oral communication.
