Independence via Joint and Marginal Distributions
What You'll Learn
You'll learn the factorization criterion: two random variables are independent if and only if their joint distribution splits into the product of their marginal distributions.
The Core Idea
Remember from the previous lesson that two random variables X and Y are independent when knowing one tells you nothing about the other. There's a beautiful mathematical test for this: check if the joint distribution factors.
For discrete random variables:
- X and Y are independent if and only if f(x,y) = f_X(x) · f_Y(y) for all possible values x and y
- Here f(x,y) is the joint PMF, and f_X(x), f_Y(y) are the marginal PMFs
For continuous random variables:
- X and Y are independent if and only if f(x,y) = f_X(x) · f_Y(y) for all x and y
- Here f(x,y) is the joint PDF, and f_X(x), f_Y(y) are the marginal PDFs
Think of it like separating ingredients: if you can write the "combined recipe" (joint distribution) as simply "ingredient A times ingredient B" (marginal distributions), then A and B don't interact—they're independent.
Why This Matters
This factorization is both a definition and a test:
- If someone hands you marginal distributions and claims variables are independent, you can multiply them to get the joint distribution
- If someone gives you a joint distribution, you can compute the marginals and check if they multiply back to give the original joint—if not, the variables are dependent
If the factorization fails for even one pair (x,y), the variables are dependent.
Key Takeaway: Independence means the joint distribution factors completely: f(x,y) = f_X(x)f_Y(y) for every x and y. This factorization is the mathematical signature of "no interaction."