Getting Started
Everything you need before the first session
Where and when
Mondays, 09:00–11:15, room MAD 099. First session: 14 September 2026.
The full session-by-session plan is in the Seminar Description.
What you need to bring
A laptop you can install software on, and that you bring to every session. This course is hands-on: you write code in almost every session, and the final exam is written on your own machine.
If you do not have a suitable laptop, contact me before the second session so we can find a solution.
The three tools
| Tool | What it is | Cost |
|---|---|---|
| R | The programming language we compute with | Free |
| RStudio | The program you actually work in | Free |
| Quarto | Turns your code and text into a finished report | Free |
All three are free and open source, on Windows, macOS and Linux.
Recommended versions: R ≥ 4.4, RStudio ≥ 2024.04, Quarto ≥ 1.5.
Please try to install all three before the first session. Step-by-step instructions are in Gräbner-Radkowitsch (2026b) and Gräbner-Radkowitsch (2026c), both on Method and Inquiry.
If something goes wrong, do not worry and do not spend hours on it — bring your laptop to session 1 and we will fix it together. But please try first, so we know where the problems are.
What you already know
You have had Research Methods / Statistik I and II. That means you have already met descriptive statistics, correlation, confidence intervals, significance tests and regression — and you have used R to prepare data.
Statistik I and II did not cover producing graphics in R. You can read a chart; you have probably never built one. Two full sessions of this course are devoted to it, and it is the skill you are most likely to use in your first job.
This course does not repeat Statistik. Where we come back to regression, it is to do something different with it — predicting, rather than explaining.
If your statistics feels rusty, work through Gräbner-Radkowitsch (2026d). It covers conceptual foundations, data types and scales, descriptive statistics in R, probability, sampling and inference, and hypothesis testing — everything this course builds on.
GitHub account
From session 4 onwards we use Git and GitHub to keep track of our work and to hand in reports 2 and 3.
You need to register at github.com — a free account is enough. Do it before session 4. Use your university email address, which also lets you claim the free GitHub Student Developer Pack.
We work inside the applied-data-science-euf organisation. To be invited, send me your GitHub username as a Moodle message — the username, not your email address. Reports 2 and 3 are then handed in through Classroom 50; the Classroom 50 manual explains every step.
Session 4 is on 19 October, and it starts from a working setup rather than building one — thirty people installing software at once is not a good use of the room.
- Work through Git and GitHub with RStudio (Gräbner-Radkowitsch 2026a): install Git, and connect RStudio to your GitHub account
install.packages("usethis")
Allow about half an hour. If something goes wrong, post in the Moodle forum before Monday — that is much better than arriving stuck.
Using AI in this course
You are allowed and encouraged to use free AI assistants (ChatGPT, Claude, Gemini, …) to help you write and debug R code. We treat this as a skill to be learned properly, not as cheating — session 7 is devoted to it, including the ways AI confidently gets statistics wrong.
Two rules apply throughout:
- You are responsible for every line. If you cannot explain what a piece of code does, or why a result means what you claim, it does not count as your work.
- Declare it. Each report ends with a short note on how you used AI.
The same applies in the exam: free tools are permitted, paid ones are not.
How a session works
The sessions usually follow a similar rhythm:
| Time | Block |
|---|---|
| 09:00–09:15 | Debrief of the take-home task (when there is one) |
| 09:15–09:45 | Input lecture — the concepts you need |
| 09:45–10:10 | Live coding — I code, you watch and ask questions |
| 10:10–10:20 | Break |
| 10:20–10:55 | In-session exercise — you work, alone or in pairs |
| 10:55–11:10 | Debrief of the exercise |
| 11:10–11:15 | Intro to the take-home task (when there is one) |
In sessions without a take-home task, the extra 20 minutes go to the lecture and the exercise.
Bring your laptop, charged, to every session. The in-session exercise is over half an hour of you writing code, every week, and there is no way to do it without a machine.
Everything you need for a session is on its page beforehand. The slides, the live coding document and the exemplary solution are published after the session has taken place.
Assessment
You must pass all three Quarto reports to be admitted to the exam. They are written individually and graded pass/fail only. Each covers the material of the sessions before it.
If a report does not pass you may revise and resubmit it. You have one revision in total across the three reports — not one per report — so use it where it matters.
Report 1 goes through Moodle. Reports 2 and 3 go through GitHub, with Classroom 50 — we set that up in session 4. The Report 1 task sheet is posted on Moodle; Reports 2 and 3 have an assignment link on the Reports page, and the task sheet comes in the repository it creates for you. The Reports page here carries the deadlines and the rules.
The final exam is 120 minutes, open book, written on your own laptop and handed in through Classroom 50, like Reports 2 and 3. Full details in the Seminar Description.
Data used in this course
| Sessions | Data | What a row is |
|---|---|---|
| 1–3 | Vienna hotels | One hotel, price for one night, November 2017 |
| 5–8 | European hotels | One hotel-date, across 46 cities |
| 9 | House sales, Ames (Iowa) | One sale, 2006–2010 |
| 9 (optional) | Airbnb listings, Hackney (London) | One listing, March 2017 |
| 10 | Telecom customers (IBM sample data, fictional) | One customer |
| 11 | Hungarian firms | One firm in 2012, with whether it exited a year later |
| 11 (exercise) | Telecom customers, as in session 10 | One customer |
All except the telecom customers and the Ames house sales come from Békés and Kézdi (2021) and are licensed CC BY-NC-SA 4.0. The Ames data is described in De Cock (2011); it and the telecom customers are taken from the R package modeldata, whose MIT licence covers the package — the data themselves are teaching datasets (the Ames assessor’s records compiled by De Cock, and IBM’s fictional telecom sample). Each session page says where its data came from and what, if anything, was changed.
Literature
There is no textbook you need to buy — the material here is self-contained. These are excellent companions, and all are free online:
- Wickham et al. (2023) — the main reference for R and the tidyverse
- Békés and Kézdi (2021) — where most of our data comes from, with the business context
- Healy (2018) — for the visualization sessions
- Ismay and Kim (2020) — a gentle R-based introduction to visualization and inference
- James et al. (2021) — for prediction, regularization and the machine learning sessions
Where things happen
| Channel | Used for |
|---|---|
| This website | All course material, session by session |
| Moodle course 17916 | Announcements, questions, submissions, grades — enrolment password AdvData26 |
| GitHub | Reports 2 and 3, from session 4 |
Stuck?
- The Tutorials page collects everything you need for installation and your first steps in R
- Post your question in the Moodle forum — that way everyone benefits from the answer
- Or make an appointment during my office hours