Bringing it together

Research Methods — Week 10

Recap

The full journey

You’ve designed an investigation, collected and analysed data, checked assumptions, fitted models, and identified limitations.

Now: write it up so someone else can understand it and act on it.

• [5 min] This is the last content delivery. It should leave students confident about what a good report looks like.

Questions?

Submit questions anonymously:

PollEv.com/geol

text geol to 07480 781235

• Open the Poll Everywhere Q&A activity before the session and leave it active throughout.
• Questions appear in your Poll Everywhere presenter view; upvoted questions rise to the top.
• Address during transitions or at the end of concept blocks, not mid-explanation.
• Reset (archive) the activity between sessions so questions don’t carry over from the previous week.

Report structure

🎓 Concept block 1

The 4-page policy report

Section	Length	Purpose
Summary	½ page	The finding, the evidence, the recommendation
Introduction	½ page	The question, why it matters
Methods	½–1 page	Data, approach, why these tests
Results	1–1½ pages	Figures, tests, plain-language interpretation
Discussion	1 page	What it means, caveats, recommendation

• [20 min block] Walk through each section using the Phase 1 faithful exemplar as a concrete example.

The summary test (revisited)

A reader who stops after the summary should still know:

1. What you investigated
2. What you found
3. What you recommend

If any of these are missing, the summary needs work.

Methods: enough to reproduce

Your methods section should answer:

• Where did the data come from?
• What did you do to clean/prepare it?
• Which tests, and why?
• What software and versions?

Could someone clone your repo and replicate your analysis?

Results: figures that tell a story

Every figure should have:

• A clear “so what”
• A caption that tells the reader what to see
• Axis labels with units
• A reference in the text

Every test result should include:

• The test used
• The p-value and an effect size or CI
• A plain-language interpretation

Summary swap

✏️💬 Exercise 1

Trade summaries

Swap your draft summary with someone from a different group.

As a reader, answer:

1. What is the main finding?
2. Do I trust it?
3. What would I do with this information?

If the reader can’t answer these from the summary alone, it needs work.

• [15 min] Students write (if they haven’t) or refine, then swap and discuss.

Peer review in science

🎓 Concept block 2

How it works

Submit → Editor assigns reviewers → Anonymous critique → Revise and resubmit.

• [20 min block] Peer review as self-correction mechanism.

Why it matters

• It’s how science self-corrects
• Catches errors the authors missed
• Challenges assumptions
• Improves clarity

Its weaknesses

• Slow
• Biased toward established researchers and ideas
• Doesn’t catch fraud well
• Can be superficial

What makes a good review

• Specific — not “this could be better” but “the CI is missing from Table 2”
• Constructive — what to fix, not just what’s wrong
• Focused on argument and evidence — not just grammar

Your Week 4 GitHub Issues review was peer review. You’ll do it again now, with higher standards.

Common pitfalls

💬✏️ Exercise 2: HolmesCo’s greatest hits

Quick-fire: spot the problem

Six excerpts from fictional reports. Two minutes each.

• [15 min] These are cumulative — each pitfall connects to a specific week’s content. Students should recognise most of them.

Pitfall 1

“The data clearly demonstrate that offshore wind is the most cost-effective energy source.”

Problem: Conclusion not supported by the data shown. The report only compared three sources and used levelised cost from 2019.

Pitfall 2

A figure with no caption, no axis labels, and no reference in the text.

Problem: The reader can’t interpret it. A figure that isn’t discussed is decoration, not evidence.

Pitfall 3

“p = 0.04, therefore the effect is significant and the policy should be adopted.”

Problem: No effect size, no CI, no discussion of practical significance. “Is that a big number?”

Pitfall 4

“The results prove that solar power will eliminate fossil fuel use by 2040.”

Problem: Overstatement. “Prove” is almost never appropriate. Extrapolation without caveats.

Pitfall 5

“We used statistical analysis to test the hypothesis.”

Problem: Methods too vague to reproduce. Which test? Why? On what data?

Pitfall 6

A report with no limitations section.

Problem: Every analysis has limitations. Omitting them doesn’t make them disappear — it makes the reader distrust you.

Telling the story honestly

🎓 Concept block 3

The key habits

These are the questions that run through the whole course:

• [20 min block] Tie the whole arc together. Each refrain came from a specific week.

The refrains

“Is that a big number?” / “Compared to what?”

Week 2

“What assumptions are we making?”

Week 3

“How plausible was this before we tested?”

Week 6

“What is the model not capturing?”

Week 8

The analyst you want to be

A good report uses these questions reflexively.

A HolmesCo report ignores them.

Your goal: be the analyst whose work a policy-maker can trust.

• This is the emotional capstone. Students have learned these habits through 10 weeks of practice. Name them explicitly and show that each came from a specific moment in the course.

Peer review briefing

📋 For the application session

The process

In the application session:

1. Finalise your draft (20 min)
2. Review someone else’s report from a different group (40 min)
3. File 3–4 GitHub Issues using the updated checklist
4. Debrief and revise (30 min)

The review template is more demanding than Week 4’s — it now includes effect sizes, assumptions, reproducibility, and overstatement checks.

• Distribute review assignments now or at the start of the application session.

Wrap-up

Next time

Application session: “Review, revise, submit”

Bring a complete draft. Whatever is committed at minute 30 gets reviewed.

This is the last session. Make it count.

• The stakes are clear. Students should arrive with a near-complete draft.