I am a quantitative research scientist and evaluation consultant who turns messy, real-world data into trustworthy evidence people can use. My work spans education, public health, and applied research, from data extraction and quality review through statistical modeling, interpretation, and stakeholder delivery.
| Area | Evidence in this portfolio |
|---|---|
| Data analysis and BI | SQL metrics, dimensional modeling, Power BI-ready measures, operational dashboards, data-quality checks, and decision-ready reporting |
| Data science and research | Quasi-experimental designs, propensity-score methods, multilevel models, diagnostics, and robustness checks |
| Reproducible delivery | Documented assumptions, synthetic data generation, automated tests, GitHub Actions, audit trails, and project controls |
| Project | What it demonstrates |
|---|---|
| Student success operations dashboard | End-to-end SQL and Power BI-ready BI project with a star schema, DAX, three dashboard views, tested metrics, implementation-risk prioritization, and an executive decision memo |
| SQL analytics case study | SQL data modeling, CTEs, window functions, cohort retention, anomaly review, tested outputs, and a decision memo |
| Student success predictive modeling | Temporal validation, probability calibration, capacity-aware thresholding, subgroup diagnostics, responsible-use controls, and reproducible R scoring |
| Administrative data pipeline | Stata and R workflows that standardize, deduplicate, join, audit, and test messy multisource administrative records |
| Quasi-experimental program evaluation | Propensity-score matching, covariate-balance diagnostics, clustered inference, robustness checks, and parallel R/Stata workflows |
| Multilevel outcomes analysis | Three-level longitudinal data, mixed-effects modeling, variance decomposition, diagnostics, and interpretation |
| Structural equation modeling | Confirmatory factor analysis, measurement invariance, FIML, latent-variable mediation, model diagnostics, and reproducible R/lavaan testing |
| Evaluation data-quality toolkit | Data contracts, automated validation, test coverage, audit reporting, and reusable SQL checks |
| Research project-management toolkit | Charters, work plans, risk and decision controls, stage gates, change management, and evaluation governance |
Methods: program evaluation, propensity-score methods, hierarchical linear modeling, mixed methods, measurement and assessment analysis, statistical reporting
Core analytic tools: Stata and R
Data and reporting: SQL, AWS Athena, Power BI
Additional exposure: Tableau and Snowflake
Reproducible workflows: Git, GitHub, automated tests, and continuous integration
Project management: PMP certification expected August 2026
All portfolio data are synthetic. The repositories are designed to make the full workflow shareable—including assumptions, quality checks, code, tests, outputs, and interpretation—without exposing client or participant information.