Skip to content

Teacher guide — Module 6 (1 period)

Goal: apply Grade 8 statistics to a real data set the students produced themselves, and practise the critical point that activity ≠ quality.

Timing: hook 5 · collect 10 · chart 15 · interpret 5 · exit 5. "Understand it" can be read at home before the lesson. Materials: graph paper, rulers, coloured pencils; printed practice-repos/g08-class-magazine/log-sample.txt for Tier 0.

Common errors

  • Counting commit lines rather than commits per date. Model grouping on the board: circle equal dates, then tally.
  • Scale choices that make bars unreadable; ask pairs to justify their scale.
  • Reading the chart as a measure of quality. Use exit question 3 to discuss: commits measure activity, not quality.
  • --date=short missing → long dates that are hard to group. Check the command.
  • On Windows sort | uniq -c works in Git Bash, not in PowerShell — use Git Bash for the bonus.

Fallbacks

Tier 0: use the printed log-sample.txt (dates are included); Tier 1: LibreOffice Calc optional.

Tell the mathematics teacher: this lesson practises frequency tables, bar charts, mode and mean with self-generated data. A joint lesson works well.

Differentiation

  • Support: give a pre-drawn axis frame.
  • Stretch: exercises 1 and 3; ask "what would a histogram of commit sizes (lines changed, from git log --stat) look like?"

Answer key

Answers depend on the repo. For log-sample.txt: 2026-10-01 → 2 commits, 10-02 → 1, 10-05 → 2, 10-07 → 1 (total 6). Busiest days: 10-01 and 10-05 (two modes). Mean over the 4 working days = 6 ÷ 4 = 1.5 commits per day. Exit 3: disagree — messages and content quality matter; many tiny meaningless commits are not better. Free-resource 1: bar chart — authors are categories and we compare counts; a pie works only if parts of a whole matter and there are few authors. Free-resource 2: e.g. git log --format="%h %s" → short id and subject.

Extension quest

Chart "commits per hour of the day" with --format=%ad --date=format:%H.