Statistics for HCI. Alan Dix

Скачать книгу

Maths and more

       1.2 Do you need stats at all?

       1.3 The job of statistics –from the real world to measurement and back again

       1.3.1 The ‘real’ world

       1.3.2 There and back again

       1.3.3 Noise and randomness

       1.4 Why are you doing it?8

       1.4.1 Empirical research

       1.4.2 Software development

       1.4.3 Parallels

       1.5 What’s next

       PART I Wild and Wide –Concerning Randomness and Distributions

       2 The unexpected wildness of random

       2.1 Experiments in randomness

       2.1.1 Rainfall in Gheisra

       2.1.2 Two-horse races

       2.1.3 Lessons

       2.2 Quick (and dirty!) tip

       2.2.1 Case 1 –small proportions

       2.2.2 Case 2 –large majority

       2.2.3 Case 3 –middling

       2.2.4 Why does this work?

       2.2.5 More important than the math

       2.3 Probability can be hard –from goats to DNA

       2.3.1 The Monty Hall Problem

       2.3.2 Tip: make the numbers extreme

       2.3.3 DNA evidence

       3 Properties of randomness

       3.1 Bias and variability

       3.1.1 Bias

       3.1.2 Bias vs. variability

       3.2 Independence and non-independence

       3.2.1 Independence of measurements

       3.2.2 Independence of factor effects

       3.2.3 Independence of sample composition

       3.3 Play!

       3.3.1 Virtual two-horse races

       3.3.2 More (virtual) coin tossing

       3.3.3 Fair and biased coins

       3.3.4 No longer independent

       4 Characterising the random through probability distributions

       4.1 Types of probability distribution

       4.1.1 Continuous or discrete?

       4.1.2 Finite or unbounded

       4.1.3 UK income distribution –a long tail

       4.1.4 One tail or two?

       4.2 Normal or not?

       4.2.1 Approximations

       4.2.2 The central limit theorem –(nearly) everything is Normal

       4.2.3 Non-Normal –what can go wrong?

       4.2.4 Power law distributions

       4.2.5

Скачать книгу