Project 4: Final Manuscript & Presentation

NHANES

Project Overview

This final project brings together everything you have learned about R, data science, and applied public-health analytics. You will produce a scientific manuscript written entirely in Quarto, using NHANES data to answer a meaningful health research question.

Your manuscript will include:

  • a clearly defined research question
  • a transparent analytic workflow
  • descriptive data summaries
  • regression modeling
  • inference and uncertainty
  • thoughtful interpretation grounded in public health

You will also give a 10-minute presentation summarizing the study and findings in a professional, conference-style format.

This project is designed to simulate real-world research writing in an MPH context.

Dataset

You will continue using the NHANES dataset from the {NHANES} package. You may refine your analytic dataset based on feedback from Projects 2 and 3.

Learning Objectives

By completing this project, you will be able to:

  • write a scientific manuscript using reproducible workflows
  • communicate data analysis and modeling in clear language
  • interpret results in a biomedical and equity-aware context
  • produce publication-quality tables and figures
  • present findings orally in a professional manner

Manuscript Structure

Your manuscript must follow this structure.

Title Page

Include:

  • title
  • author name
  • course and institution
  • date

Abstract

Write 150 to 250 words summarizing:

  • background
  • purpose
  • methods
  • key findings
  • implications

Introduction

Explain:

  • the health topic
  • why it matters
  • key background or epidemiologic context
  • your research question(s) or hypothesis

Write for an informed public-health audience.

Methods

Describe:

  • data source and sample
  • variables and coding
  • missing-data handling
  • analytic plan, including models used
  • ethical considerations

Be transparent and precise.

Results

Include:

  • descriptive statistics
  • at least one regression model
  • tables and figures produced in R
  • uncertainty measures such as confidence intervals
  • plain-language interpretation

Let the data speak clearly.

Discussion

Address:

  • key findings
  • possible mechanisms or interpretation
  • equity implications
  • limitations
  • directions for future research
  • practice or policy relevance

Avoid overstating causality.

References

Use APA format.

Appendix

Include code excerpts or supplemental tables as needed.

Presentation

You will give a 10-minute presentation including:

  • research question
  • methods overview
  • key results
  • interpretation
  • reflection

Slides should be clear and professional.

Formatting Requirements

Your manuscript must:

  • be written completely in Quarto
  • knit without errors
  • use inline code for statistics where possible
  • include at least two figures and one table
  • be approximately 2,000 to 3,000 words, excluding code
  • reflect professional writing quality

Submission

Submit:

  • your knitted manuscript in word
  • your .qmd source file
  • your presentation slides

Academic Integrity

All writing and code must be your own. Collaboration on ideas is permitted; copying is not.

Project 4 Rubric

1. Research Question and Framing — 15 points

Evaluates clarity, significance, and grounding.

  • 13–15: Clear, meaningful health question; strong rationale; excellent framing
  • 10–12: Good question with moderate depth
  • 6–9: Question vague or weakly justified
  • 0–5: Lacks clarity or purpose

2. Methods and Transparency — 15 points

Assesses documentation and rigor.

  • 13–15: Methods clearly described; coding choices transparent; reproducible workflow obvious
  • 10–12: Mostly clear with some gaps
  • 6–9: Limited clarity or detail
  • 0–5: Methods unclear or inappropriate

3. Data Analysis and Modeling — 20 points

Evaluates execution and appropriateness.

  • 17–20: Models appropriate; data handling correct; output interpreted properly
  • 13–16: Minor errors but overall sound
  • 8–12: Substantial misunderstanding or weaknesses
  • 0–7: Incorrect or incomplete analysis

4. Interpretation, Uncertainty, and Public-Health Insight — 20 points

Focuses on meaning-making.

  • 17–20: Thoughtful interpretation with nuance; uncertainty addressed; equity implications considered
  • 13–16: Interpretation generally correct with moderate depth
  • 8–12: Limited interpretation
  • 0–7: Incorrect or missing interpretation

5. Tables, Figures, and Presentation of Results — 10 points

Assesses clarity and professionalism.

  • 9–10: Publication-quality outputs; clearly labeled; visually effective
  • 7–8: Mostly clear but could be improved
  • 4–6: Limited polish or clarity
  • 0–3: Poor or missing presentation

6. Writing Quality and Organization — 10 points

Evaluates readability and tone.

  • 9–10: Clear, cohesive, professional writing throughout
  • 7–8: Mostly clear with minor issues
  • 4–6: Some difficulty following structure
  • 0–3: Poor clarity or organization

7. Reproducibility and Quarto Workflow — 5 points

Rewards transparent scientific computing.

  • 5: Fully reproducible; clean structure; knits without errors
  • 3–4: Minor issues present
  • 1–2: Inconsistent reproducibility
  • 0: Not reproducible

8. Oral Presentation — 5 points

Evaluates clarity and professionalism.

  • 5: Engaging, clear, organized, and confident
  • 3–4: Generally clear with minor gaps
  • 1–2: Difficult to follow
  • 0: Presentation not delivered