Welcome, Recap & Tooling Upgrade

Session 1 — 12 March 2026

Prof. Dr. Claudius Gräbner-Radkowitsch

Europa-Universität Flensburg

2026-03-12

Logistics

What is this course?

Building on what you know: R, Quarto, basic linear regression — all from Research Methodology (Fall 2025)

What we add:

  • Regression done properly: multiple predictors, binary outcomes, diagnostics
  • Causal thinking: why correlation ≠ causation, and what to do about it
  • Panel data, modern tools, professional communication
  • Learn to use tools such as GitHub, which are essential for today’s working environment

Guiding idea: Every session starts with a business question

  • Not: “Today we learn logistic regression”
  • But: “Which firms are likely to exit the market — and can we predict it?”

Full course schedule


# Date Title
1 Thu, 12 Mar Welcome, recap & tooling upgrade ← today
— 19 + 26 Mar Take-home task 1
2 Thu, 09 Apr Multiple regression: going beyond the basics
— 16 + 23 Apr Take-home task 2
3 Thu, 30 Apr Modelling binary and categorical outcomes
4 Thu, 07 May What can go wrong: biases and diagnostics
— Thu, 14 May No lecture
5 Thu, 21 May Causation vs. correlation: thinking like an economist
6 Thu, 28 May Panel data and fixed effects
7 Thu, 04 Jun Coding smarter: R and AI tools
8 Thu, 11 Jun Communicating data: advanced visualization
9 Thu, 18 Jun Recap and looking ahead

Assessment

Take-home tasks (mandatory, pass/fail)

  • Two structured analyses submitted as rendered Quarto reports (task 1 as html, task 2 as pdf)
  • Assigned during breaks — self-organize and use the discussion space
  • Submit via GitHub Classroom
Task Due Topic
Task 1 02 April, 23:59 Regression basics (recap)
Task 2 29 April, 23:59 Extending a regression

Final exam

  • Open-book, on-site, individual
  • Details confirmed later in the semester

The break weeks

19 and 16 March, 16 and 23 April

This time is intentional and structured:

  • Work on Take-home Task 1 and 2 at your own pace
  • Review material from the prior course if anything feels rusty
  • Post questions in GitHub Discussions — I will check regularly

Tip

If something in Task 1 feels unfamiliar, that is a signal — not a problem. The two weeks are designed for exactly this.

How this course works


Where you code: GitHub Codespaces — RStudio in your browser, zero installation

  • Alternative: install R Studio and packages locally (as in the previous course)

    • Advantage: faster, works offline, more independence
    • Disadvantage: potential challenges during installation
    • See tutorials from previous course

Where you ask questions: GitHub Discussions on the course repository (not email, not Moodle — so everyone benefits from the answer)

Where you submit: GitHub Classroom

What you need:

  • A GitHub account (free)
  • GitHub Student Developer Pack (recommended — free professional tools)

Today’s plan


Time Block
16:00–16:20 Welcome, course overview, assessment
16:20–17:00 Quarto and Git upgrade: live demo
17:00–17:15 Questions and break
17:15–18:15 Student exercise
18:15–19:00 Introduce Task 1, Q&A


The exercise today is about workflow setup, not data analysis.

→ No dataset required — we are building a Quarto template for later

Quarto and Git(Hub)

Quarto in this course

What Quarto is:

  • A next-generation scientific publishing system
  • Write narrative text, R code, and output (tables, figures) in a single .qmd document

Why Quarto matters:

  • Code and text in one place — no copy-pasting numbers into Word
  • Results update automatically when data or code changes
  • Renders to HTML, PDF, Word — same source file
  • Professional standard for reproducible, shareable data analysis in academia and industry

The standard we set today applies to every submission:

  • Proper front matter, bibliography, clean and well-readable output

Git and GitHub

  • Git: a version control system: tacks changes you make to your files
    • Always see who did what when - and roll back if needed
  • GitHub: stores the Git info on a cloud and facilitates publishing and collaboration
    • A platform of great importance - more than ever in the AI era

Git: key concepts

  • Repository (‘repo’): a folder for which you tack changes via Git
    • Fork (of a repo): your personal copy of someone else’s repo
  • Key Git actions:
    • Stage: Mark changes to include into next snapshot
    • Commit: Save named snapshot of all staged changes
    • Revert: Roll back to the state of a previous commit
  • Additional actions if Github is involved:
    • Push: Send your local changes to GitHub
    • Fetch: Check what has changed on Github
    • Pull: Fetch and apply remote changes to your local repo

Git: key concepts

  • These five actions are essential.
    • Branching and merging is ‘only’ super useful
  • All can be done through the RStudio Git pane — no command line necessary.

Tip

For a full introduction: GitHub Skills — Introduction to GitHub

Your coding environment: two options

Option A — Fork + Codespace

Codespaces are a cloud-based coding environment with everything pre-installed

  • No installation problems, but only works in the cloud and slower than local installation

Fork the course repository → your own copy on GitHub → open Codespace from your fork

  • In the VS Code terminal, type rserver → login: username rstudio, password rstudio
  • You can save and commit your own work freely
  • Next session: open your existing Codespace at github.com/codespaces — do not create a new one
  • When new materials arrive: Sync fork on GitHub → Pull in RStudio

Option B — Local installation

Install R 4.4+, RStudio, and the required packages (see course website)

Tip

First thing to do today: go to the course repository and click Fork

Quarto upgrade: live demo document

  • I will develop a Quarto document as a demo
  • We will also use it to demonstrate how GitHub classroom works
  • You do not need to make any notes, everything will be made available
  • You can do the test for GitHub classroom yourself at home using the link from the website

Q&A and Break

Student Exercise

Student Exercise


Tip

Register at GitHub if not already one

  1. Go to the website
  2. Accept the Classroom student example exercise
  3. Do the exercise
  4. Commit and push your solution in the end

Ask questions whenever they occur

After 18:15 we clarify remaining issues

Notes on take-home task 1

Take-home task 1

What determines weekly earnings of market research analysts in the United States?

Dataset: U.S. Current Population Survey (CPS MORG), 2014 — market research analysts and marketing specialists

Four tasks:

  1. Clean the data and produce a descriptive statistics table
  2. Simple regression: earnings ~ age — visualise and interpret
  3. Quadratic extension: add age² — does fit improve? Where do earnings peak?
  4. Add controls: hours worked and college degree — discuss omitted variable bias

Deadline: 02 April 2026, 23:59 Submit: GitHub Classroom (link is on website)

See you on 9 April

Next session: Thu, 09 April — Multiple regression: going beyond the basics

Questions during the break? GitHub Discussions on the course repository

Task 1 deadline: 02 April 2026, 23:59

Good luck — and welcome to the course.