Introduction & overview

Theoretical and Empirical Research Methodology · Implementation Lab 1

Prof. Dr. Claudius Gräbner-Radkowitsch

Europa-University Flensburg, Department of Pluralist Economics

18 September 2026

Part I: Organization & outlook

The goal of the implementation lab

Prepare, analyse and present quantitative data with R — implementing the tools of the main lecture in practice

Why R?

Among the most widely used languages for statistical analysis
Demanded in almost every industry
The whole pipeline in one consistent, reproducible framework
Free, open source, OS-independent → inclusive and transparent
A gateway to other programming languages
A large and friendly community

The days of commercial statistical languages and packages such as SAS, Stata and SPSS are over Paul Jansen, CEO of Tiobe Software

#	TIOBE Sep ’26	PYPL Aug ’26	RedMonk Q1 ’26
1	Python	Python	JavaScript
2	C	Java	Python
3	C++	C/C++	Java
4	Java	R	PHP · C#
5	C#	JavaScript	–
6	JavaScript	Objective‑C	TypeScript
7	Visual Basic	PHP	…
8	SQL	…	…
9	R	…	…
10	Rust	…	…
13			R

TIOBE: search results · PYPL: tutorial searches · RedMonk: GitHub + Stack Overflow

TIOBE: R climbed #15 → #9 within a year

What you will be able to do

Read in data from many sources
Prepare messy data → produce tidy data

country,1952,1957,1962,1967,1972,1977,...
Afghanistan,Asia|28.801|8425333|779.445,...
Albania,Europe|55.23|1282697|1601.056,...

↓

# A tibble: 142 × 5
   country       continent lifeExp     pop gdpPercap
   <fct>         <fct>       <dbl>   <int>     <dbl>
 1 China         Asia         73.0  1.32e9     4959.
 2 India         Asia         64.7  1.11e9     2452.
 3 United States Americas     78.2  3.01e8    42952.
 # … with 139 more rows

What you will be able to do

Create publication-ready visualisations
Identify hidden patterns in data and make predictions using models

What you will be able to do

Write reproducible reports in Quarto
Publish them on the web via Netlify
Acquire tools that later seminars build on

one source file → a finished paper

But can’t ChatGPT just do this for us…?

Yes and no 😇
LLMs are genuinely useful when developing code
You know the language → LLMs boost your productivity
You don’t know the language → LLMs confuse you
- Slightly misbehaving code you cannot spot
- Blind to the most recent developments in the language
- Output is not always easy to read

We will learn to use LLMs for coding — they are allowed throughout, including in the exam — but be wary of their limitations 🤓

Organization of the lab sessions

Every on-site session = theory + practice → always bring your laptop 👩‍💻
- Some topics work better as videos → watch, do the exercises, bring your questions
Questions belong in the Moodle forum, not in emails
- Try to answer each other — explaining is how you learn
- Recorded once, available to everybody
- The best questions open the next session

A short tour through the website

researchmethodology-fall26.netlify.app

Always reload the page — you want the most recent version

Our ‘learning agreement’

Why do I expect regular activity from you?

Learning to program is consecutive → miss the basics, get lost later
This is a demanding course → catching up later is hard
It works through practice → the exercises carry a huge benefit
It is at times frustrating → we need an amicable environment
Few things teach you more than helping someone else

Learning a programming language can be a lot of fun and really brings you forward — if we do it together as a team 💪

Open questions?

Part II: Installing R and R Studio

R and R-Studio

R — the programming language

Issues commands your computer understands

> fib_n(4)

↓

8B542408 83FA0077 06B80000 0000C383
FA027706 B8010000 00C353BB 01000000
B9010000 008D0419 83FA0376 078BD989
C14AEBF1 5BC3

↓

[1] 3

R Studio — the IDE

A fancy text editor with features that make programming easy

R and R-Studio

R = the engine

R Studio = the dashboard

Figure: Ismay & Kim (2022)

Install R first, then R Studio
Afterwards you only ever open R Studio → it calls R when needed

R and R packages

Base R understands a lot of commands — but not everything
You can teach R new routines by defining functions
You can save those functions and pass them on to others

An R package = a collection of functions and data written by others, installed once and then usable like anything built into R

R and R packages

R = a new phone

Packages = the apps you download

Figure: Ismay & Kim (2022)

A script installs every package we need this semester — so we can resolve all installation issues right now

And what about LaTeX?

In Quarto: R code and text in one file → HTML page or polished PDF
HTML is easy — a PDF is not: it is produced by LaTeX, a typesetting system that turns plain text into properly set text

---
title: "R Markdown and PDF"
author: "Claudius"
format: pdf
---

Here I write text. And add code:

```{r}
sqrt(1:5)
```

Quarto source …

… becomes LaTeX …

… becomes a PDF

Installation procedure

Start today, not the night before

Trying to install R just before the next session is a guaranteed way to lose it.

Until next session, you should have…
- …tried to install R, R Studio and Git → follow the tutorials
- …posted every problem with a screenshot in the Moodle forum
All installation problems must be solved by the end of next week

Problems with the installation?

Check the tutorials again
Post your problem in Moodle
Always attach a screenshot