AcademyProbability

Academy

Expectation

Level 1 - Math II (Physics) topic page in Probability.

Principle

Expectation is a probability-weighted average. It describes the long-run centre of a random quantity, not necessarily a value that can occur in one trial.

Notation

\(X,Y\)
random variables
\(E[X]\)
expectation, or mean, of X
\(p_X(x)\)
PMF of a discrete random variable
\(f_X(x)\)
PDF of a continuous random variable
\(g(X)\)
a function of the random variable X
\(a,b\)
constants

Method

Step 1: Multiply each value by its probability or density weight

For a discrete variable, add value times probability. For a continuous variable, integrate value times density.

Discrete expectation
\[E[X]=\sum_{x\in D_X}x\,p_X(x)\]
Continuous expectation
\[E[X]=\int_{-\infty}^{\infty}x f_X(x)\,dx\]

Step 2: Average functions by applying the function first

If the quantity of interest is \(g(X)\), weight \(g(x)\), not \(x\).

Discrete function
\[E[g(X)]=\sum_{x\in D_X}g(x)p_X(x)\]
Continuous function
\[E[g(X)]=\int_{-\infty}^{\infty}g(x)f_X(x)\,dx\]

Step 3: Use expectation rules to simplify

Linearity lets constants pass through expectation and lets sums be averaged term by term.

Rules

Discrete mean
\[E[X]=\sum_{x\in D_X}x p_X(x)\]
Continuous mean
\[E[X]=\int_{-\infty}^{\infty}x f_X(x)\,dx\]
Function of a variable
\[E[g(X)]=\sum_{x\in D_X}g(x)p_X(x)\]
Linearity
\[E[aX+b]=aE[X]+b\]
Additivity
\[E[X+Y]=E[X]+E[Y]\]
Positivity
\[X\ge0\quad\Rightarrow\quad E[X]\ge0\]
Independent product
\[X,Y\ \text{independent}\quad\Rightarrow\quad E[XY]=E[X]E[Y]\]

Examples

Question
Find the expected score on a fair die.
Answer
Each value has probability
\[1/6\]
\[E[X]=1\cdot\frac{1}{6}+2\cdot\frac{1}{6}+3\cdot\frac{1}{6}+4\cdot\frac{1}{6}+5\cdot\frac{1}{6}+6\cdot\frac{1}{6}=\frac{21}{6}=3.5.\]

Checks

  • Expectation has the same unit as the random variable.
  • An expected value need not be a possible observed value.
  • Probabilities or densities must define a valid distribution before computing expectation.
  • Additivity does not require independence.
  • The product rule \(E[XY]=E[X]E[Y]\) does require independence in this course setting.