Joint probability and cumulative distribution functions (discrete)

Welcome to the World of Pairs!

Up until now, you have likely spent most of your time looking at one random variable at a time (like the number of claims in a day). But in the real world—and on Exam P—things rarely happen in isolation. Usually, we are interested in how two or more variables behave together. This is what we call Multivariate Random Variables.

In this chapter, we focus on Discrete variables. Think of this as looking at a "grid" of possibilities rather than a single line. Don't worry if this seems tricky at first; once you see the patterns in the tables, it becomes much easier!

1. The Joint Probability Mass Function (PMF)

When we have two discrete random variables, \(X\) and \(Y\), we want to know the probability that both happen at the same time. We write this as \(p(x, y)\) or \(P(X = x, Y = y)\).

The "Restaurant Order" Analogy

Imagine you are at a fast-food place. \(X\) is the number of burgers a customer orders (0, 1, or 2), and \(Y\) is the number of sodas (0, 1, or 2). The Joint PMF tells you the probability of a specific combination, like "What is the chance a customer orders exactly 2 burgers and 1 soda?"

The Rules of Joint PMFs

Just like with single variables, there are two golden rules for a valid joint PMF:
1. Non-negativity: Every single probability must be 0 or greater: \(p(x, y) \ge 0\).
2. The Grand Total: If you sum up every single cell in your grid/table, the total must equal 1: \(\sum_x \sum_y p(x, y) = 1\).

Quick Review: The Table View

Most Exam P questions for discrete joint variables will give you a table. The rows represent values of \(X\), and the columns represent values of \(Y\). Each cell is the joint probability for that specific pair.

Key Takeaway: The joint PMF is just a way to map out the "and" probabilities for two variables at once.

2. Marginal Probabilities: "Summing Out"

Sometimes, we have a table for \(X\) and \(Y\), but we suddenly only care about \(X\). This is called finding the Marginal Distribution.

How to do it

If you want the marginal probability of \(X\), you sum across all possible values of \(Y\). It’s like ignoring the soda orders and just looking at the total burger count.

Formula for \(X\): \(p_X(x) = \sum_{all y} p(x, y)\)
Formula for \(Y\): \(p_Y(y) = \sum_{all x} p(x, y)\)

Memory Aid: Why is it called "Marginal"?

Think of a piece of paper. The margins are the edges. To find the marginal probability, you sum the rows or columns and write the totals in the edges (margins) of your table!

Common Mistake to Avoid: When calculating the marginal for \(X\), students sometimes accidentally sum the \(X\) values. Remember: To find the "total" for a specific \(X\), you must add up all the \(Y\) values that could have happened with it.

3. The Joint Cumulative Distribution Function (CDF)

The Joint CDF, written as \(F(x, y)\), is the probability that \(X\) is less than or equal to a certain value AND \(Y\) is less than or equal to a certain value.

\(F(x, y) = P(X \le x, Y \le y)\)

The "Bottom-Left Corner" Rule

In a discrete table, if you want to find \(F(1, 2)\), you simply find the cell at \((1, 2)\) and add up every probability that is to the left of it and below it (including itself). It’s like highlighting a rectangle starting from the origin up to your point \((x, y)\).

Did you know?

The Joint CDF always has these properties:
- It starts at 0 (if \(x\) or \(y\) are very small).
- It ends at 1 (if \(x\) and \(y\) are at their maximum possible values).
- It is non-decreasing. As you increase \(x\) or \(y\), the probability can only stay the same or go up.

4. Step-by-Step: Solving a Discrete Joint Problem

If you encounter a problem with a joint PMF formula like \(p(x, y) = k(x + y)\) for \(x = 1, 2\) and \(y = 1, 2\):

Step 1: Create a table. List out all possible pairs: (1,1), (1,2), (2,1), (2,2).
Step 2: Plug in the values. (e.g., for (1,1), the value is \(k(1+1) = 2k\)).
Step 3: Solve for \(k\). Sum all the cells and set them equal to 1. (\(2k + 3k + 3k + 4k = 12k\), so \(k = 1/12\)).
Step 4: Answer the question. If they ask for the marginal \(P(X=1)\), just sum the row where \(X=1\).

Summary and Key Takeaways

1. Joint PMF is the probability of \(X\) and \(Y\) happening together: \(P(X=x, Y=y)\).
2. Marginal PMF is found by "summing out" the variable you don't want. Totaling a row gives you the marginal for that row's variable.
3. Joint CDF is the "accumulated" probability: \(P(X \le x, Y \le y)\).
4. Tables are your friends! For discrete variables, drawing a grid is usually the fastest and safest way to avoid calculation errors.

Don't worry if the double summations look scary! Just think of them as adding up numbers in a box. You've got this!