Applied Replication for Data Skills
A hands-on workshop · Open Research Conference, Newcastle University
This workshop aims to introduce several technical skills involved in modern empirical data analysis – retrieving data from an online repository, working with computer code in a reproducible fashion, version-controlling your analysis and publishing a reproducible report – through an applied “replication” exercise.
Participants will reproduce one of the reanalyses that were part of the Multi100 project, whose results were recently published in Nature (Aczel et al. 2026), and test the robustness of a published claim from Teney (2016) – using the same data, tools, and workflow the Multi100 project used.
Event details
| Event | Open Research Conference, Newcastle University |
| Date | Tuesday 16 June 2026 |
| Time | 10:15–12:45 |
| Room | Room 1.06, Henry Daysh Building |
| Programme | https://go.ncl.ac.uk/orc |
What you will do
The workshop centres on Teney (2016), a study of EU attitudes across 27 countries. You will:
- Reproduce the constrained result that five independent analysts published as part of the Multi100 project – confirming the same coefficient from the same data.
- Deviate (with reasons) – choose one justifiable alternative from a specification menu, commit to your choice before you run it, run the model, and submit a standardised result.
- See the multiverse – your result, alongside your classmates’ and the five published analysts’, on one chart.
The exercise is a miniature version of Multi100 itself: same claim, five-plus analysts, explicit choices, comparable results.
Two tracks – pick the one that fits your setup
Open the Browser lab page. The R engine runs entirely in your browser via webR – no R, Positron, or Git needed on your machine. The page will take 10–30 seconds to load the first time; read the task while it warms up.
This track covers the full analytical exercise. The only thing you cannot do in-browser is the git workflow; for that step, you will use the GitHub web interface.
You have R, Positron, Quarto, and Git installed. You will clone the template repository, work locally, commit as you go, and publish your report to GitHub Pages automatically via a pre-built workflow.
Track B is the richer experience: you leave the session with a live public page, a committed git history, and a portfolio piece.
Not set up yet? Go to Setup now – it takes some time, depending on what you need to install.
Work in pairs to help each other out. If you do not have a laptop, try pairing up with someone who has; you can still complete Track A on the classroom computer, but also get an insight on the Track B workflow.
Session schedule
| Time | Segment |
|---|---|
| 10:15 | Welcome and pairs formed |
| 10:20 | Talk I – the three Rs: reproducibility, robustness, replicability (via the three Nature 2026 papers) |
| 10:40 | Talk II – the Teney case: original claim, SCORE reproduction attempt, five-analyst findings |
| 10:50 | Setup sprint: get a track working, or verify yours is ready |
| 11:05 | Concepts: estimands and DAGs – mini-lecture then guided exercise |
| 11:20 | Break |
| 11:35 | Reproduce the constrained result (t ≈ −3.80) |
| 11:45 | Your analytical path: choose a specification, preregister, run, report |
| 12:05 | Share and publish |
| 12:15 | Multiverse debrief: the class specification curve |
| 12:30 | Wrap and resources |