Plankton Research Group Symbol Plankton Research Group

Plankton Research Group · Programme under development

As-If Science Jam

Data literacy as playful practice.

The As-If Science Jam is a format experiment of the Plankton Research Group. We know what we want to show. We know which thresholds we want to lower. We know by which route we want to try — playfully, in a protected setting, with real data. What we don't yet know is the final form. That is precisely why the Jam is an open format development.

What we set out to do

Aim 01

Make it visible

The micro-techniques by which data turn into convincing yet misleading statements — cherry-picking, p-hacking, survivorship bias, axis choice, confounding, Simpson's paradox. Not explained, but played through.

Aim 02

Lower thresholds

Statistics is too often treated as territory only experts may enter. That impression paralyses public debate around numbers. The Jam reverses it: everyone gets to manipulate, everyone gets to check.

Aim 03

Playful, but serious

Data literacy is not learned in lectures but through own experience in a protected setting. The Jam is a game with serious tools — the learning curve lies in the tension between the two.

Core narrative

We live in a world that argues with models and statistics — but few engage in data-based discourse. Two toxic extremes coexist: blind faith in numbers and fear of mathematics.

The Jam addresses both. It deliberately perverts the scientific process in order to see through it. That way, art can stand beside science and make a dry process alive and accessible.

First iteration · 3-day format

This is one first concrete shape — not the final one. It will be tried first in cooperation with Schmiede Hallein. What we learn will feed the next iteration.

Day 1 · evening

The Hypothesis

  • · Hosts' opening: why data literacy, why playful, why „as if"
  • · Team forming (3–5 people plus mentor)
  • · Dataset buffet — teams browse the available sources
  • · Hypothesis session: each team formulates the most absurd hypothesis it wants to confirm in the data
  • · Ethics framing: Everything that happens here happens in real research every day. It is only made visible.

Day 2

The Forge

  • · Teams work with R, Python, AI tools on the „confirmation"
  • · Templates as entry point — nobody stares at an empty console
  • · Mentors actively counter skill gaps
  • · Mentors flag continuously: „That's p-hacking." „That's survivorship bias."
  • · Afternoon: transition into poster making. How do we sell the absurd thesis as convincingly as possible?

Day 3

Offensio / Defensio

  • · Poster session as the central closing format
  • · Offensio: other teams (and hosts) try to expose weaknesses — which data were left out? which axes manipulated?
  • · Defensio: the team defends — with humour, also with substance
  • · Reflection round: which of this happens in real newspapers, studies, policy papers?
  • · Close: from perversion back to construction — what is good data-based argumentation?

Ethics as a through-line

Ethics is not a separate session but runs through the whole event. The Offensio/Defensio makes the ethical dimension tangible: when someone exposes a trick, the importance of scientific integrity becomes felt. Fact-checks react to individual lies. The Jam inoculates against the lying-procedure.

Format development · open questions

The format described above is a first approximation. From each iteration we want to learn and develop it further. Key development questions:

Audiences
First iteration: Schmiede audience — creatives, makers, scientists. Later iterations: upper-secondary students, university students (especially non-STEM), interested citizens. Which format adaptations are needed for which group?
Duration
3 days at Schmiede, or more compact 1–2-day variants for school and university contexts? What depth is lost at which length?
Setting
Festival, classroom, university seminar, library, citizen workshop. Which frames amplify the playful approach, which undermine it?
Datasets
The biggest challenge. We need a curated collection with „absurdity potential" — datasets thematically far apart but temporally/spatially overlapping. This collection becomes part of our open-data pillar.
Mentor pool
5 teams of 4 would need 5 mentors with data-science background. Who is involved in which setting? How do we prepare mentors for the specific „manipulation-not-correctness" mindset?
AI tools
Explicit use of LLMs by teams for code generation, argumentation, visualisation? An interesting additional layer: can an AI help with cherry-picking? Spoiler: yes, very well — and that is part of the learning experience.

Crew · first iteration

Michael Hackl

Host · format architect

Direction of Plankton Research Group. Connects the playful with the methodological. Provides tooling and templates.

Methodological anchor

Quantitative social research

Brings quantitative rigour from a domain in which statistical claims have direct societal consequences. Mentor on quantitative questions.

Open science · translation

Science communication

Brings the question of how science is communicated. Bridge between laboratory and gallery.

Mentor names will be announced once commitments for the first iteration have been confirmed.

Get involved

The format isn't made by us alone. It is rebuilt with each iteration.

If you want to mentor, if you have a setting where the Jam might fit (school, university, festival, library), if you know interesting public datasets, or if you want to help develop the format conceptually — write to office@planktonresearch.org.

First-iteration cooperation: subnet and Schmiede Hallein. Curation: Michael Hackl. Methodological anchor: Plankton Research Group.