Linear Combinations of Random Variables
You'll learn: How to compute the mean and variance of linear combinations like aX + bY using expectation properties and independence.
What Is a Linear Combination?
A linear combination of random variables is simply a weighted sum, like 3X + 5Y - 2Z. You're scaling random variables by constants and adding them together. Think of it like mixing ingredients in a recipe: 3 cups of flour (X) plus 5 tablespoons of sugar (Y) minus 2 teaspoons of salt (Z).
The question is: what are the mean and variance of this mixture?
Computing the Mean
The mean of a linear combination is straightforward thanks to the linearity of expectation:
E[aX + bY] = aE[X] + bE[Y]
This works regardless of whether X and Y are independent! The constants pull out, and expectations distribute over addition.
Example: If X has mean 10 and Y has mean 20, then E[3X + 5Y] = 3(10) + 5(20) = 30 + 100 = 130.
Computing the Variance
Variance is trickier because it involves squared terms. For independent random variables:
Var(aX + bY) = a²Var(X) + b²Var(Y)
Notice the constants are squared and there's no cross term (because independence means covariance is zero).
Example: If Var(X) = 4 and Var(Y) = 9, then Var(3X + 5Y) = 3²(4) + 5²(9) = 36 + 225 = 261.
If X and Y are not independent, you must include the covariance term:
Var(aX + bY) = a²Var(X) + b²Var(Y) + 2abCov(X,Y)
The Big Picture
Linear combinations preserve structure beautifully. You're building new random variables from old ones, and the formulas tell you exactly how uncertainty propagates through the combination.
Key Takeaway: For linear combinations aX + bY, the mean is aE[X] + bE[Y] (always), and the variance is a²Var(X) + b²Var(Y) when independent (add 2abCov(X,Y) if not).