Seminar focus: debate what counts as clean data
Present a shared messy dataset with duplicates, blanks, and ambiguous categories. Ākonga propose different cleaning rules, test their effects, and explain which decisions are mathematical and which require contextual judgement.
- Seminar move: Competing cleaning proposals and consequence comparison
- Seminar outcome: A defensible cleaning rule with acknowledged trade-offs
🎯 Learning Intentions
- Clean data by removing errors or unclear responses
- Sort data into categories
- Use frequency tables to count responses
- Introduction to digital spreadsheets (optional)
🎥 Media Anchor (8 mins)
Video: Research Skills for Students
- Which organisation method makes pattern-finding easiest?
- How can poor data organisation distort your conclusions?
1. The Messy Desk Metaphor (5 mins)
Discuss: "Why is it hard to find a specific paper on a messy desk?"
Raw data is like a messy desk. Organising it helps us find the answers.
2. Activity: Cleaning Data (10 mins)
Look at your data set. Are there any issues?
- Did someone write "dog" when you asked for a number?
- Did someone answer twice?
- Are there blank spaces?
Task: Fix clear errors or decide to remove "spoiled" data entries.
3. Skill: Frequency Tables (20 mins)
Turn a list into a count.
Raw List:
Red, Blue, Red, Green, Blue, Red...
Frequency Table:
- Red: ||| (3)
- Blue: || (2)
- Green: | (1)
Task: Create frequency tables for your own survey questions.
4. Introduction to Google Sheets/Excel (15 mins - Optional)
Demonstrate how to enter data into a spreadsheet:
- One row per person
- One column per question
- Using "Sort" to group answers
📋 Teacher Planning Snapshot
Ngā Whāinga Ako — Learning Intentions
The class organises one dataset several ways and argues about which grouping is honest.
Ngā Paearu Angitū — Success Criteria
- ✅ I can show how different groupings change what a dataset appears to say.
- ✅ I can defend a grouping choice against a specific objection.
Differentiation & Inclusion
Scaffold support: The same dataset supplied pre-grouped three different ways. Extension: construct a grouping that is technically true and materially misleading.
ELL / ESOL: Physical card-sorting before any discussion of categories.
Inclusion: Sorting physically is a full route into the argument, not a lesser one.
Mātauranga Māori lens: How you categorise decides what you can see. Ngā tohu o te taiao are chosen signs; the choosing is itself interpretation.
Prior knowledge: Students should have basic familiarity with data displays (bar graphs, dot plots). No prior statistical investigation experience required — the PPDAC inquiry cycle provides accessible scaffolding for first-time investigators.
Curriculum alignment
This lesson develops the Level 4 statistical-investigation objective through a real stage of the PPDAC cycle. See the unit curriculum companion for the exact source statement and lesson-to-evidence map.