LECTURE HUB
Lecture 2 - Describing data without lying
🚦 Status🌟 Mastered
🧠 Mastery🦅 Adept
📅 Last Review11 May 2025
🎯KEY TAKEAWAYS
- Describe any dataset with three things: centre, spread, shape. One number is never a description.
- When the shape is lopsided, report the 🔤 Median and the middle half. The mean is telling you about the tail.
- A picture of all the points beats any summary of them. Rosling's whole talk is that sentence, animated.
- Assignment 1 forbids an average on its own. Good.
⚡QUICK JOTS
- three questions for any pile of numbers: where is the middle, how wide, what shape
- mean = balance point. median = the one in the middle of the queue
- an average on its own is a rumour (Hale)
- range is two numbers and both are the weird ones. use the middle half instead
- skew: long tail to the right → mean gets dragged right, median stays
- 20:05 he put on a talk: a Swedish doctor, bubbles moving across a chart, a whole room laughing at fertility data. 🎬 The best stats you've ever seen
- ten-minute quiz at the end. 8 of 10. lost two on quartiles
☑️HOMEWORK
- Watch the Rosling talk again at home
- Write 🔤 Median in my own words
- Pick the dataset for 📌 Statistics I - Assignment 1 - Describing a dataset
- Find out what my own dashboards report: mean or median
- Redo the two quartile questions from 📌 Statistics I - Quiz 1 - Describing data