Combining Random Variables

When we model real-world situations with probability, we rarely deal with just one uncertain quantity at a time. A delivery company tracks both the weight of a package and its travel distance; a student adds quiz scores together for a final grade; a casino computes the difference between money wagered and money paid out. In each case, we are combining random variables — forming new random variables by adding, subtracting, or otherwise operating on existing ones. Understanding how the mean and variance behave under these operations is one of the most practically powerful ideas in probability. Two foundational rules govern this behaviour. First, the mean (expected value) of a sum or difference of random variables always equals the sum or difference of the individual means — no matter what. Second, if two random variables are independent, the variance of their sum or difference equals the sum of their individual variances (note: always a sum, even for differences). These rules extend naturally to any finite collection of random variables and underpin everything from quality-control engineering to statistical inference. This article derives and explains both rules, illustrates them with worked examples, and connects them to the broader framework of probability distributions. Mastering these ideas will prepare you for studying sampling distributions, confidence intervals, and hypothesis testing in later units.

Key takeaways
  • What 'combining' means
  • The Addition Rule for Means (always true)
  • The Addition Rule for Variances (requires independence)
  • Scaling a Random Variable
  • General Linear Combination

What 'combining' means

If X and Y are random variables defined on the same probability space, we can create a new random variable W = aX + bY, where a and b are constants. For example, if X is the score on a maths test and Y is the score on an English test, then W = X + Y is the total score. The new variable W is itself random — it has its own probability distribution, mean, and variance.

The Addition Rule for Means (always true)

The expected value operator E is linear, meaning E[aX + bY] = a·E[X] + b·E[Y]. In particular, E[X + Y] = E[X] + E[Y] and E[X − Y] = E[X] − E[Y]. This holds whether or not X and Y are independent, and whether or not they have the same distribution. The proof follows directly from the definition of expectation: E[X + Y] = Σ(x+y)·P(X=x, Y=y) = Σx·P(X=x) + Σy·P(Y=y) = E[X] + E[Y].

The Addition Rule for Variances (requires independence)

Variance is not generally linear. For W = X + Y, we compute Var(W) = E[(W − μ_W)²]. Expanding gives Var(X + Y) = Var(X) + Var(Y) + 2·Cov(X, Y), where Cov(X, Y) = E[(X − μ_X)(Y − μ_Y)] is the covariance. When X and Y are independent, Cov(X, Y) = 0, so Var(X + Y) = Var(X) + Var(Y). Crucially, for W = X − Y: Var(X − Y) = Var(X) + (−1)²·Var(Y) = Var(X) + Var(Y). Variance always adds, even for differences. In standard deviation terms, SD(X ± Y) = √(Var(X) + Var(Y)) when independent.

Scaling a Random Variable

If W = aX for a constant a, then E[aX] = a·E[X] and Var(aX) = a²·Var(X). Notice the square on the constant for variance. Standard deviation scales linearly: SD(aX) = |a|·SD(X). This is why multiplying every value by 2 doubles the mean but quadruples the variance.

General Linear Combination

Combining both rules: for independent X and Y and constants a, b, c, the random variable W = aX + bY + c satisfies E[W] = a·E[X] + b·E[Y] + c and Var(W) = a²·Var(X) + b²·Var(Y). Adding a constant c shifts the mean but does not change the variance. This generalises to any number of independent random variables: E[X₁ + X₂ + … + Xₙ] = E[X₁] + … + E[Xₙ] and Var(X₁ + … + Xₙ) = Var(X₁) + … + Var(Xₙ).

The Key Equation

μX±Y=μX±μYσX±Y2=σX2+σY2(X,Y independent)\mu_{X \pm Y} = \mu_X \pm \mu_Y \qquad \sigma^2_{X \pm Y} = \sigma^2_X + \sigma^2_Y \quad (X, Y \text{ independent})
  • \mu_X — Mean of X (same units as X): The expected value (long-run average) of random variable X
  • \mu_Y — Mean of Y (same units as Y): The expected value (long-run average) of random variable Y
  • \sigma^2_X — Variance of X (units of X squared): The average squared deviation of X from its mean, measuring spread
  • \sigma^2_Y — Variance of Y (units of Y squared): The average squared deviation of Y from its mean, measuring spread
  • \sigma^2_{X \pm Y} — Variance of the combined variable (units squared): Equals the SUM of variances whether adding or subtracting X and Y, assuming independence

Worked Example

Problem. A factory produces bolts (length X, in mm) with mean μ_X = 50 mm and standard deviation σ_X = 0.4 mm, and nuts (inner diameter Y, in mm) with mean μ_Y = 50.5 mm and standard deviation σ_Y = 0.3 mm. A bolt fits inside a nut if Y − X > 0 (clearance is positive). Find the mean and standard deviation of the clearance D = Y − X, assuming X and Y are independent.

μD=0.5 mm,σD=σX2+σY2=0.16+0.09=0.5 mm\mu_D = 0.5 \text{ mm}, \quad \sigma_D = \sqrt{\sigma^2_X + \sigma^2_Y} = \sqrt{0.16 + 0.09} = 0.5 \text{ mm}

Answer. Mean clearance = 0.5 mm; Standard deviation of clearance = 0.5 mm

Figure
Diagram
Combining Random Variables: Mean and VarianceXμ_X = 50σ_X = 0.4Yμ_Y = 50.5σ_Y = 0.3−D = Y − Xμ_D = 0.5 mmσ_D = 0.5 mmMean Rule (always):μ_D = μ_Y − μ_XVariance Rule (if indep.):σ²_D = σ²_X + σ²_Y
Origins · 1 minPafnuty Chebyshev and how this was discovered
PC
Pafnuty Chebyshev
1821–1894 · Russian

Chebyshev made foundational contributions to probability theory, including the celebrated Chebyshev Inequality which places universal bounds on how spread out any probability distribution can be from its mean. His rigorous work on expected values and variance established the analytical framework that underpins the rules for combining random variables used today. Chebyshev was also a gifted mechanical engineer who designed linkages for converting rotary motion into straight-line motion — his probabilistic thinking and mechanical precision went hand in hand.

  • Expected Value of a Random Variable — Expected value (mean) is the foundation for the addition rule for means; you must understand E[X] before combining random variables.
  • Variance and Standard Deviation — Variance is the quantity that adds when combining independent random variables; knowing how variance is defined clarifies why it always adds.
  • Independence of Random Variables — The variance addition rule only holds for independent random variables; understanding independence is essential for applying these formulas correctly.
  • Normal Distribution — The sum of independent normal random variables is also normal, making the combination rules especially powerful when working with bell-curve distributions.
  • Sampling Distributions — The sampling distribution of the sample mean relies directly on combining n independent random variables, applying both the mean and variance rules.