Lecture 1

Syllabus and Course Outline

Byeong-Hak Choe

bchoe@geneseo.edu

SUNY Geneseo

August 24, 2026

One question connects the semester

A semester-long environmental-policy study

Ask	Build	Analyze	Interpret
Frame a consequential question	Find, document, transform, and audit evidence	Visualize, map, model, and validate	Explain what the evidence means—and what it cannot establish

Note

The course produces one coherent study, not a collection of disconnected exercises.

Invite each student to name an environmental issue or policy problem they might want to understand. Keep the opening broad; feasibility comes later.

How the course works

Two tracks, one research paper

Technical track

• Transform and document environmental data in R
• Design explanatory graphics
• Map and analyze spatial patterns
• Fit and validate a question-appropriate model

Research track

• Refine the research question
• Engage literature and policy debates
• Make defensible design choices
• Interpret findings, limitations, and implications

Note

Every analytical choice should strengthen the policy argument; every policy claim should be traceable to evidence.

Emphasize the recurring feedback loop between research design and code. The methods sequence can adapt to the two projects.

What you should be able to do

Frame the study

• Formulate a focused research question
• Locate the relevant literature and debate
• Choose a feasible unit, place, and time span
• Develop a defensible analysis plan

Build the evidence

• Create a reproducible data workflow
• Design effective figures and maps
• Evaluate and interpret a suitable ML analysis
• Synthesize results, limits, and policy implications

Connect these outcomes to the students’ prior preparation. Ask what feels familiar and what is new.

Visualize first; model second

Step	Purpose
Inspect	Units, coverage, missingness, provenance
Visualize + map	Patterns, anomalies, scale, uncertainty
Set a baseline	A transparent comparison
Add one extension	ML only when it serves the question
Validate + interpret	Generalization and policy limits

Tools follow the project: direct model-specific R—not tidymodels; sf by default; terra only for raster data.

Clarify that this is not a compressed DANL 310 plus DANL 320. Selective reuse is driven by the research question. Explain that terra becomes worthwhile for land cover, elevation, satellite, or climate rasters.

Assessment

Five course requirements

Requirement	What it establishes	Timing
Proposal	Define the question, evidence, and design	Set with instructor
Kick-off report	Prove feasibility with real data	Oct. 20
Progress report	Show evidence, obstacles, and next steps	Set with instructor
Final exam	Synthesize and apply course methods	Dec. 9–15; TBD
Research paper	Integrate design, analysis, and interpretation	Set with instructor

No midterm. Final-exam details and remaining dates will be confirmed through Brightspace.

Point students to Brightspace for dates or weights not yet fixed. The website and syllabus are working documents for this individualized course.

Research Kick-off Report

Due Tuesday, October 20

One week after Fall Break

The report should show:

• a refined question and clear policy stakes;
• documented sources, units, coverage, license, and provenance;
• reproducible data intake plus a quality audit;
• one question-led figure and a preliminary map when relevant;
• a baseline, justified extension, validation design, and metric—or a reason not to use ML; and
• principal risks and the next four weeks of work.

Stress that this is a scope and feasibility checkpoint, not a demand for polished results.

Four lenses for evaluation

Technical implementation in R

Transformation, visualization, spatial analysis, and machine learning are effective and reproducible.

Grasp of the research question

The work engages relevant literature and key environmental-policy debates.

Research paper

Design, analysis, writing, and policy interpretation form one integrated argument.

Comprehensive exam

Methods are synthesized, selected, applied, and interpreted appropriately.

Use one example to show that sophisticated code cannot compensate for a vague question, and a compelling topic cannot compensate for undocumented analysis.

Individualized does not mean informal

The path may adapt

• question and policy setting;
• data availability and scale;
• spatial or temporal structure; and
• current technical needs.

The standards do not

• evidence;
• documentation;
• interpretation; and
• intellectual honesty.

Class: Mon + Wed, 12:30–1:45 PM · Office hours: Mon + Wed, 1:45–3:15 PM · South 227B

Explain how individualized meetings work and where updates will be communicated. Confirm that the office is both the classroom and office-hours location.

Begin the research record

Week 1: Question + Data Scan

By the end of the week, prepare a brief working artifact with:

1. two or three candidate research questions;
2. one preferred direction and its policy stakes;
3. two candidate datasets with source links;
4. likely unit, time span, geography, and provenance; and
5. one initial plot or data-quality screenshot with the limitation it reveals.

Note

This is formative—not a new graded requirement. It creates the raw material for the proposal and the research website.

Ask students to record uncertainty rather than hide it. A failed data source is useful evidence if its limitation is documented.

Why begin with a Quarto website?

Component	Role in the research record
Local Quarto files	Keep writing, code, figures, and sources together
Git	Record meaningful snapshots of the project
GitHub	Store and share the repository and its history
GitHub Pages	Make the rendered research site reviewable

Distinguish Quarto, Git, GitHub, and GitHub Pages. They solve different problems but form one publishing workflow.

Two students, two setup paths

New to GitHub

1. Create the account and record the username.
2. Confirm Git and set name/email.
3. Store a GitHub credential securely.
4. Create USERNAME.github.io.
5. Clone it as an RStudio project.
6. Render, commit, push, and verify.

Website already exists

1. Open the correct local project.
2. Pull and confirm a clean starting state.
3. Audit _quarto.yml and file organization.
4. Add or refine a research page.
5. Render the full site.
6. Commit, push, and verify.

Warning

A personal access token is a password. Paste it only into a credential prompt—never into a script, .qmd file, screenshot, or Git commit.

Let the experienced student work through the audit path while you guide the novice through account and credential setup. Rejoin at the shared publishing cycle.

One verified publishing cycle

Edit → render → inspect → commit → push → verify

quarto render
git add .
git commit -m "Add the first research update"
git push

Before considering the setup complete, verify both:

• the repository at https://github.com/USERNAME/USERNAME.github.io; and
• the public site at https://USERNAME.github.io.

Note

Each student should finish with one successful, visible, and reproducible update.

Open the course Git, GitHub & Quarto guide

Review the syllabus · Open the Week 1 hub

Keep the last slide visible during setup. Confirm the local render before the first commit, then verify both GitHub URLs after pushing.