Appendix A. Semester Capstone Project
This appendix describes the course-long capstone project. You build one analysis across the whole semester, one phase at a time, so that by the final week you have a complete, auditable study of a healthcare question you chose yourself.
Work on the project in phases. You complete the phases over the course of the semester, one at a time, as each matching topic is covered in class. Each phase builds on the one before it, and you may revise earlier phases as your understanding grows. Keep everything in one well-organized workbook and project folder, so the whole project stays traceable from question to answer.
The instructions below are high level on purpose. Your instructor will provide the specific due dates, deliverable format, and grading rubric for each phase.
Phase 1. Question and stakeholders
Alongside asking good questions and defining measures.
- Choose a healthcare question that genuinely interests you.
- Write a short paragraph explaining why it matters.
- Identify two different audiences or stakeholders, and describe what each one cares about in your question.
- Sketch your initial thoughts on what data you would ideally collect to answer it.
Phase 2. Data source and ethics
After data structures and healthcare data governance are covered.
Find a real, publicly available data source for your question. Describe its variables and state the unit of analysis. Break your main question into two or three sub-questions. Identify the ethical issues in how the data was collected and how it may be used. Revise Phase 1 if your source changes what is feasible.
Phase 3. Data preparation and codebook
After planning data collection and data cleaning are covered.
Assemble your data into a tidy, spreadsheet-ready table. Write a codebook that defines every column so someone else could use your data correctly. Clean and organize the data, and document the steps. Discuss the population your data represents and whether it is a sample or the full population.
Phase 4. Summaries and visuals
After descriptive statistics and PivotTables are covered.
Produce summary statistics and frequency counts for your key variables. Interpret what they say in plain language. Create at least two visuals and interpret each one. Focus on what the numbers mean for your stakeholders, not just what they are.
Phase 5. Models and answer
After correlation and regression are covered.
Examine relationships in your data with correlations and at least two scatterplots with trend lines. Build at least two PivotTables that explore your question. Interpret the scatterplots and the PivotTables, state any predictions your models support, and finally answer your original research question. Revise earlier phases as needed so the finished project reads as one coherent study.
Phase 6. Presentation
Once your analysis is complete.
Present your finished project to a small group: the question, the data, your summaries, your visuals, your findings, and your next steps. Practice explaining your analysis to a non-technical audience, leading with what matters.