14.3. Marginal Distributions

To study individual distributions from a joint distribution, we use marginal distributions.

Marginal distributions allow us to focus on the distribution of one variable while ignoring the other in a joint probability scenario. Specifically, when we have a joint distribution of two random variables, such as X and Y, the marginal distribution lets us analyze either X or Y independently. This is useful because it simplifies statistical analysis and makes it easier to understand the individual behavior of each variable.

Imagine that you are in a large classroom with students from different majors. If you want to understand how students in general perform in mathematics, you could look at the average scores of each major separately (marginal score) rather than considering their scores collectively (joint distribution). This way, you gain insights about each major's performance independently.

Marginal PMF (Discrete)

• Marginal PMF of 𝑋: $P (x) = \sum P(X = x, Y = y)$
• Marginal PMF of 𝑌: $P (y) = \sum P(X = x, Y = y)$

In the context of discrete random variables, the marginal probability mass function (PMF) is calculated by summing the joint probabilities over the other variable. For example, to find the marginal PMF of X, we sum the probabilities of all pairs that include the specific value of X while varying Y. Similarly, the marginal PMF of Y is found by summing over all possible values of X. This enables us to determine the probabilities associated with X and Y independently.

If you were tracking the weather conditions in a city each day for a month, and you recorded both temperature (X) and humidity (Y), the marginal PMF for temperature would show you the probabilities of experiencing specific temperatures without worrying about what the humidity was on those days.

Marginal PDF (Continuous)

• Marginal PDF of 𝑋: $f (x) = \int_{-\infty}^{\infty} f (x, y) dy$
• Marginal PDF of 𝑌: $f (y) = \int_{-\infty}^{\infty} f (x, y) dx$

In the case of continuous random variables, we define the marginal probability density function (PDF) by integrating the joint PDF over the entire range of the other variable. For instance, to find the marginal PDF of X, you integrate the joint PDF 'f' over all possible values of Y. The same process applies for obtaining the marginal PDF of Y by integrating over all values of X. This technique helps us determine the likelihood of each of the random variables independently from the joint distribution.

Imagine a study measuring the heights and weights of individuals in a large population. To find the marginal PDF of heights, you would integrate the joint distribution of height and weight across all weights, effectively showing how heights alone are distributed among the population without considering their weights.

--