Teacher guide — Grade 11, Unit 8: Reproducibility, Git and the climate project¶

Time: 3 lessons + project (about 5 periods). Track: B (project).

#	Lesson	Time
1	Git basics	40–45 min
2	Seeds, environments, notes	40–45 min
3	Project: analyse and publish	40–45 min
P	Unit project: Climate or weather?	2 periods

Before you start¶

Materials: Notebook notebooks/climate_project.ipynb; Git installed (https://git-scm.com).
Read docs/pedagogy/lesson-anatomy.md and docs/pedagogy/ai-use-policy.md.

If you have no computers / no internet¶

Practise 'commit messages' on paper by describing each change in one line.

Common misconceptions¶

Saving as final_v2_REAL_final.ipynb is version control.
Setting a seed makes a result true; it only makes it repeatable.

Differentiation¶

Support: give a partially filled table / sentence starters; pair with a data-steward role.
Stretch: ask "what would change your conclusion?" and "who is missing?"

Answer key (check questions)¶

Lesson 1: Git basics 1. What does git commit record?
Answer: A snapshot with a message. 2. Why write clear messages?
Answer: So you know what and why later. 3. Which command shows changes waiting?
Answer: git status.

Lesson 2: Seeds, environments, notes 1. Why set a seed?
Answer: So random results can be repeated. 2. What else affects reproducibility?
Answer: Library/Python versions and data versions. 3. Does a seed make a result correct?
Answer: No: only repeatable.

Lesson 3: Project: analyse and publish 1. What proves it is reproducible?
Answer: Someone else reruns it and gets the same results. 2. What goes in the limits?
Answer: Synthetic data, one series, short period, weather vs climate. 3. What do you disclose?
Answer: AI help used, data licence and source.

Assessment¶

Starter quiz each lesson (3 questions, one recalling the previous lesson); unit quiz in quiz.yml; project marked with the rubric in docs/pedagogy/assessment.md; 2-minute oral check per group.

Alignment¶

Enrichment aligned to NEB Grade 11–12 Computer Science (databases, programming) and Maths/Statistics.

Answers: practice exercises¶

Edit the file → git status → git add → git commit.
"Fix typo in rainfall chart title"; "Add 95% bootstrap interval for temperature trend"; "Remove raw CSV from repo and add data/raw/ to .gitignore" (also note it remains in history; ask the teacher how to purge if sensitive).
Creates a random-number generator with seed 42, so the same "random" numbers come out each run.
data/raw/ .ipynb_checkpoints/ secrets.txt
No seed; cells run out of order; data file changed (downloaded again); different library versions; hidden state from deleted cells still in memory.
Put the file in the project (data/climate.csv) and use pd.read_csv("data/climate.csv") (or Path(__file__)- based paths in scripts).
Seeds 1–5 (1,000 resamples each): 0.18–0.39, 0.19–0.39, 0.18–0.39, 0.17–0.39, 0.18–0.38 °C/decade — the conclusion (positive trend, interval excludes 0) is robust.
It proves the notebook works from a clean start, with no hidden variables from cells you deleted or ran out of order.
Missing code/data, undocumented manual steps, different software versions, errors, selective reporting. Trust: if others cannot check a result, policy based on it rests on faith.
Check for: question, data source/licence/date, how to run, seed/versions, key results, limits, AI use, licence.

Real-data exercises: notes¶

R1 hashlib.sha256(open(path, "rb").read()).hexdigest(); teach the difference between "same code" and "same data".
R2 OWID and WDI revise estimates each year; small differences are normal — record the download date.
R3 Supervise any public contribution; follow CONTRIBUTING.md and the code of conduct.