Is expected value the same as the mean?

Q: Which symbol should I use on the AP exam?

Use μ_X or E(X) when the question hands you a probability distribution, and x-bar when it hands you data. If you define your symbols in words the first time you use them, a grader can follow your work even if your notation is unusual.

By Jude Wallis · Published July 31, 2026

Yes and no: they are the same idea in two settings. Expected value is the mean of a probability distribution, found by weighting each value by its probability. The sample mean is the average of data you actually collected. The sample mean estimates the expected value.

AP Statistics: Unit 2 (topics 2.9 Parameters of Random Variables, 1.7 Summary Statistics for One Quantitative Variable). The expected value of a discrete random variable is Unit 2 topic 2.9 of the Fall 2026 AP Statistics course, which calls it the mean of the random variable and asks for an interpretation in long-run terms. The sample mean is introduced in Unit 1 topic 1.7 as a summary statistic for one quantitative variable.

Same idea, two different settings

Both are averages. They differ in what they average.

Expected value is the mean of a probability distribution. You have no data yet. You have a list of possible values and the probability of each, and you weight every value by its probability. The symbol is $E(X)$ , read "the expected value of X", or $\mu_X$ , read "mu sub X".

The sample mean is the average of data you actually collected. You add the observed values and divide by how many there are. The symbol is $\bar{x}$ , read "x-bar".

So the expected value is what the model says the average will be in the long run. The sample mean is what your data says it was this time. When a question says "the mean of the random variable X", it means the expected value; the two phrases are interchangeable.

The notation: E(X), mu sub X, and x-bar

Most of the confusion here is a symbol problem, not a concept problem.

Symbol	Read as	Averages what	Fixed or varying
$E(X)$ or $\mu_X$	expected value of X, mu sub X	every possible value of $X$ , weighted by its probability	fixed
$\bar{x}$	x-bar	the values in one particular sample	varies from sample to sample
$\mu$	mu	every value in a whole population	fixed

The pattern is the one the whole course runs on: Greek letters ( $\mu$ , $\mu_X$ , $\sigma$ ) label parameters, which are fixed numbers describing a model or a population, and Latin letters ( $\bar{x}$ , $s$ ) label statistics, which are computed from a sample and change when the sample changes. If that split is shaky, read parameter vs statistic before going on.

One warning about $E$ . It is a function applied to a random variable, so you write $E(X)$ , never $E$ on its own and never $E(x)$ with a lowercase $x$ , which would name a single observed value rather than the variable.

The two formulas have the same shape

The expected value of a discrete random variable multiplies each value by its probability and adds the products:

\mu_X = E(X) = \sum x_i \cdot P(x_i)

The sample mean adds the observed values and divides by the count:

\bar{x} = \frac{\sum x_i}{n}

Those look different, but they are the same operation. Dividing by $n$ is the same as giving every observation the weight $1/n$ , so

\bar{x} = \sum x_i \cdot \frac{1}{n}

and if several observations repeat, you can group them and weight each distinct value by its relative frequency instead. Both formulas are "value times weight, summed". Expected value uses probabilities as the weights. The sample mean uses relative frequencies. That is the entire difference, and it is why the two numbers land near each other when the sample is large: relative frequencies settle down toward probabilities.

The sample mean estimates the expected value

This is the working relationship between the two, and it is what makes inference possible.

The sample mean is an unbiased estimator of the underlying mean, written $\mu$ for a population and $\mu_X$ for a random variable, which are the same quantity under two names: if you took every possible sample of size $n$ and averaged all those sample means, you would get $\mu$ exactly. In symbols, the sampling distribution of $\bar{x}$ is centered at $\mu$ , with standard deviation $\sigma / \sqrt{n}$ when the observations are independent, which in practice means sampling with replacement or a sample no more than 10% of the population. Larger samples give sample means packed more tightly around the expected value, which is the law of large numbers doing its work.

So the arrow points one way. The expected value is the target. The sample mean is the estimate that aims at it and misses by a random amount you can measure. Everything in Units 3 and 4 is built on that sentence. For the shape and spread of that estimate, see sampling distributions explained.

Where the two genuinely differ

Three differences matter in practice.

One is fixed, one moves. $\mu_X$ is a single number set by the probability model. $\bar{x}$ is a different number for every sample you draw. Calling a sample mean "the expected value" throws away that distinction and is the fastest way to lose a point on an AP interpretation.
You can compute an expected value with zero data. A probability distribution is enough. A sample mean needs observations.
Neither has to be a value you could observe. A fair die has $E(X) = 3.5$ , which no face shows, and a sample mean can be just as unreachable as a single outcome: two rolls of 1 and 2 give $\bar{x} = 1.5$ , a value no individual roll can produce. The long-run-average reading is covered in can expected value be impossible.

A fourth difference is quieter. The standard deviation that goes with $\mu_X$ uses probability weights and divides by nothing, while the sample standard deviation $s$ divides by $n - 1$ . Do not mix the two formulas.

How the AP exam phrases it

Topic 2.9 of the Fall 2026 course is called Parameters of Random Variables, and it treats "the mean of a random variable" and "the expected value" as the same quantity. Expect all three of these wordings for the same task:

"Find the expected value of $X$ ."
"Find the mean of the random variable $X$ ."
"Find $\mu_X$ ."

Interpretation is graded separately from calculation, and the phrasing that earns the point is long-run language in context: "over many repetitions, the average number of goals per game will be about 1.1". A statement like "we expect 1.1 goals next game" is the version that loses credit, because it reads as a prediction for one trial. You can check your arithmetic with the expected value calculator and drill the interpretations in expected value practice.

Expected value from a probability distribution

Let $X$ be the number of goals a team scores in a game, with the distribution below. Find $\mu_X$ and interpret it.

$x$	0	1	2	3
$P(x)$	0.3	0.4	0.2	0.1

Check that this is a legal probability distribution first: $0.3 + 0.4 + 0.2 + 0.1 = 1$ , and every probability is between 0 and 1.
Multiply each value by its probability: $0 \times 0.3 = 0$ , $1 \times 0.4 = 0.4$ , $2 \times 0.2 = 0.4$ , $3 \times 0.1 = 0.3$ .
Add the four products: $0 + 0.4 + 0.4 + 0.3 = 1.1$ .
So $\mu_X = E(X) = 1.1$ goals.
For the spread, use the probability-weighted squared deviations: $(0 - 1.1)^2(0.3) = 0.363$ , $(1 - 1.1)^2(0.4) = 0.004$ , $(2 - 1.1)^2(0.2) = 0.162$ , $(3 - 1.1)^2(0.1) = 0.361$ .
Add them for the variance: $0.363 + 0.004 + 0.162 + 0.361 = 0.89$ , so $\sigma_X = \sqrt{0.89} = 0.9434$ goals.

$\mu_X = 1.1$ goals with $\sigma_X = 0.9434$ goals. Over many games, this team averages about 1.1 goals per game, and a typical game differs from that average by about 0.94 goals. Note that 1.1 is not a score any single game can produce.

The sample mean of data from that same model

The team plays 12 games and scores 0, 1, 1, 0, 2, 1, 3, 0, 1, 2, 1, 0 goals. Find $\bar{x}$ , and compare it with the $\mu_X = 1.1$ from the model above.

Add the 12 observations: $0 + 1 + 1 + 0 + 2 + 1 + 3 + 0 + 1 + 2 + 1 + 0 = 12$ .
Divide by $n = 12$ : $\bar{x} = \frac{12}{12} = 1.0$ goal.
Now redo it as a weighted average to see the link. The counts are four 0s, five 1s, two 2s, and one 3, so the relative frequencies are $4/12 = 0.3333$ , $5/12 = 0.4167$ , $2/12 = 0.1667$ , and $1/12 = 0.0833$ .
Weight each value by its relative frequency: $0(0.3333) + 1(0.4167) + 2(0.1667) + 3(0.0833) = 0 + 0.4167 + 0.3334 + 0.2499 = 1.0$ , the same answer.
Compare the weights with the model probabilities: the sample gave 0.3333, 0.4167, 0.1667, 0.0833 where the model says 0.3, 0.4, 0.2, 0.1. Close, but not equal, which is exactly why $\bar{x}$ and $\mu_X$ differ.
The gap is $1.0 - 1.1 = -0.1$ goals, ordinary sampling variability for only 12 games.

$\bar{x} = 1.0$ goal against $\mu_X = 1.1$ goals. The sample mean is a statistic that estimates the fixed parameter $\mu_X$ ; it came out 0.1 below the target here and would land somewhere else after the next 12 games.

Frequently asked questions

Is expected value just another name for the mean?

It is another name for the mean of a random variable or probability distribution, which is why $\mu_X$ and $E(X)$ label the same number. It is not another name for the sample mean $\bar{x}$ , which is computed from data and changes from sample to sample.

Does the sample mean ever equal the expected value exactly?

Sometimes, by luck. There is no guarantee, and with a continuous variable it essentially never happens. What is guaranteed is that $\bar{x}$ is unbiased: its sampling distribution is centered at $\mu$ , and its spread, $\sigma / \sqrt{n}$ for independent observations, shrinks as $n$ grows.

Which symbol should I use on the AP exam?

Use $\mu_X$ or $E(X)$ when the question hands you a probability distribution, and $\bar{x}$ when it hands you data. If you define your symbols in words the first time you use them, a grader can follow your work even if your notation is unusual.

Can expected value and sample mean use the same formula?

Yes, in weighted form. Both are the sum of value times weight. Expected value uses probabilities as weights; the sample mean uses relative frequencies, or equivalently the weight $1/n$ on every observation. The formulas differ only in where the weights come from.

Is the expected value a parameter or a statistic?

A parameter. It describes a model or a population and does not depend on any sample you drew. The sample mean is a statistic. That split is why one gets a Greek letter and the other gets a Latin one.