Class set · a printable, ages 15–18 and teachers
Journal club in four roles
Pick one paper a month. Start with one figure, not the abstract. Four people, four jobs; swap roles
next time. Keep the Pitfall Card on the table.
Describer
Says what the figure shows, before anyone says what it means.
- What is on each axis, or in each panel, and in what units?
- What kind of picture is it: an experiment, a prediction or a toy model?
- What does the eye land on first, and is it what the caption says matters?
Notes:
Sceptic
Finds the strongest reason the figure might not mean what the authors say.
- How many things were measured (what is n), and how were they chosen?
- What would the figure look like if the claim were wrong?
- Which square on the Pitfall Card could apply here?
Notes:
Translator
Puts the finding into plain words for someone outside the room.
- Say the finding in two sentences with no jargon.
- Which word in the paper needed a gloss, and what is it?
- What should a news story about this figure not say?
Notes:
Auditor
Checks what the paper gives so others can check it.
- Are the data and code available, and where?
- Is the confidence of each prediction shown, and how?
- Is the paper free to read, and under what licence?
Notes:
How a session runs (about 40 minutes)
- Everyone looks at the one figure for three minutes, without reading the text.
- The Describer goes first. Nobody interprets until the figure has been described.
- The Sceptic, then the Translator, then the Auditor. Each has five minutes.
- Open discussion. End by writing one sentence the group agrees the figure shows, and one it does not.
Paper: ________________________________ Figure: ______ Date: ____________
Three open-access papers to start with
- Jumper J, Evans R, Pritzel A, et al. (2021). Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589. doi:10.1038/s41586-021-03819-2. Open access, CC BY 4.0. Start with Figure 1.
- Tunyasuvunakool K, Adler J, Wu Z, et al. (2021). Highly accurate protein structure prediction for the human proteome. Nature 596, 590–596. doi:10.1038/s41586-021-03828-1. Open access, CC BY 4.0. Start with Figure 1.
- Varadi M, Anyango S, Deshpande M, et al. (2022). AlphaFold Protein Structure Database: massively expanding the structural coverage of protein-sequence space with high-accuracy models. Nucleic Acids Research 50, D439–D444. doi:10.1093/nar/gkab1061. Open access, CC BY 4.0. Start with Figure 1.
Each is free to read at its DOI. The four roles are a design for a discussion, not a tested
method; no study shows what this format teaches.