Loading module…

What you will learn in Statistics & Data Handling

1. Kinds of Data
Categorical, ordinal, discrete, continuous — and why the label decides which averages you are allowed to compute.
2. Frequency and Grouped Frequency Tables
Choosing a class interval is a judgement. Make it badly and the shape of the data disappears.
3. Charts and What They Are For
The chart is chosen by the kind of data and the question being asked. Every wrong chart is drawable, and this chapter draws them.
4. Measures of Central Tendency
Three averages doing three different jobs. The argument between them is the content.
5. Averages from Grouped Data
The grouped table replaces every value with its class mid-point. Here is what that costs, as a number.
6. Quartiles, Percentiles and Box Plots
Five numbers, drawn — and two very different datasets that produce almost the same picture.
7. Variance and Standard Deviation
Why the obvious measure of spread fails, why squaring fixes it, and why we then take the square root back.
8. Comparing Spread
The standard deviation cannot compare a spread in kilograms with a spread in seconds. This is the measure that can.
9. Outliers and Skewness
A rule for flagging unusual values, no rule for removing them, and a shape you can read off three averages.
10. Scatter Plots and Correlation
How to read a scatter, what r does and does not measure, and the three-word sentence everybody can recite and nobody applies.
11. The Least-Squares Line
Fit a line by hand, fail to beat the algorithm, then push the line past the data and watch it fail.
12. Sampling and Bias
Five ways of choosing whom to ask, the mechanism each one gets wrong, and why a bigger sample does not help.
13. How Graphs Mislead
Build the misleading chart yourself, then see the honest one. Nothing you do will change a single number.
14. Testing a Claim (Introduction)
Significance as a count of shuffles: how often chance alone produces an effect this large.

What you will learn in Statistics & Data Handling

1. Kinds of Data

Categorical, ordinal, discrete, continuous — and why the label decides which averages you are allowed to compute.

2. Frequency and Grouped Frequency Tables

Choosing a class interval is a judgement. Make it badly and the shape of the data disappears.

3. Charts and What They Are For

The chart is chosen by the kind of data and the question being asked. Every wrong chart is drawable, and this chapter draws them.

4. Measures of Central Tendency

Three averages doing three different jobs. The argument between them is the content.

5. Averages from Grouped Data

The grouped table replaces every value with its class mid-point. Here is what that costs, as a number.

6. Quartiles, Percentiles and Box Plots

Five numbers, drawn — and two very different datasets that produce almost the same picture.

7. Variance and Standard Deviation

Why the obvious measure of spread fails, why squaring fixes it, and why we then take the square root back.

8. Comparing Spread

The standard deviation cannot compare a spread in kilograms with a spread in seconds. This is the measure that can.

9. Outliers and Skewness

A rule for flagging unusual values, no rule for removing them, and a shape you can read off three averages.

10. Scatter Plots and Correlation

How to read a scatter, what r does and does not measure, and the three-word sentence everybody can recite and nobody applies.

11. The Least-Squares Line

Fit a line by hand, fail to beat the algorithm, then push the line past the data and watch it fail.

12. Sampling and Bias

Five ways of choosing whom to ask, the mechanism each one gets wrong, and why a bigger sample does not help.

13. How Graphs Mislead

Build the misleading chart yourself, then see the honest one. Nothing you do will change a single number.

14. Testing a Claim (Introduction)

Significance as a count of shuffles: how often chance alone produces an effect this large.