Random Variables and the Normal Distribution
A random variable attaches a number to the outcome of a random experiment, and its expected value and variance say where that number sits on average and how far it strays. This unit builds the rules for sums and multiples, then the three distributions that carry most of applied statistics: the binomial for counts of successes, the Poisson for counts of rare events, checked against the Prussian horse-kick deaths, and the normal curve, read term by term, standardized, and tabulated, with the 68-95-99.7 rule, the normal approximation to the binomial, and a first warning that heavy-tailed quantities such as incomes and city sizes are not normal.
Unit 3 assigned probabilities to events, such as “the coin lands heads” or “the patient tests positive”. Most questions in statistics are about numbers rather than about events: how many heads, how many deaths, how tall. This unit attaches a number to every outcome of a random experiment and asks what that number is on average, how far it typically strays from its average, and which shapes its distribution can take. Three shapes, the binomial, the Poisson, and the normal distribution, cover a large part of applied statistics, and the last of them, the bell curve, is the curve on which Units 5 to 8 rest.
1 Random variables and their distributions
A random variable is a number whose value is decided by the outcome of a random experiment. The number of heads in ten tosses of a coin is a random variable, and so are the sum of two dice, the height of a person drawn at random from a population, and the number of emails that arrive in the next hour. Random variables are written with capital letters, \(X\) or \(Y\), and the values they can take with small letters, so that \(\mathrm{P}(X = 3)\) is the probability that the number of heads turns out to be three.
Unit 1 divided variables into categorical and quantitative ones and the quantitative ones into discrete and continuous. The same division applies here. A discrete random variable has values that can be listed, usually whole numbers that count something. A continuous random variable, such as a height or a waiting time, can take any value in an interval, and Section 5 treats it separately, because probabilities then have to be handled as areas.
For a discrete random variable, the probability distribution is the list of its possible values together with their probabilities. The function that gives the probability of each value, \(\mathrm{P}(X = x)\), is also called the probability mass function. By the axioms of Unit 3 every entry is at least zero and all entries add up to one, since the variable has to take one of its values. A probability distribution is shown as a table or as a bar chart with one bar per value.
For example, a European roulette wheel has 37 numbered pockets, 0 to 36, of which 18 are red, 18 are black, and the zero is green. A player who puts 1 euro on red receives 2 euros back if the ball stops in a red pocket, which is the stake plus 1 euro of winnings, and nothing otherwise. The net gain \(X\) of this bet, in euros, is therefore \(+1\) with probability \(18/37 = 0.486\) and \(-1\) with probability \(19/37 = 0.514\), and these two lines are its whole probability distribution.