Welcome to the "Heart" of Probability!

In your journey toward passing Exam P, you have already learned how to find probabilities. But as an actuary, you don't just want to know if something *might* happen; you want to know what happens "on average" and how "spread out" the risks are. This chapter covers Expected Values, Moments, and Percentiles—the tools we use to summarize entire distributions into a few meaningful numbers.

Don't worry if these terms sound a bit mathematical at first. Think of this chapter as learning how to describe a person: the Mean is their average height, the Variance is how much their height differs from others, and Percentiles tell us where they stand in a crowd. Let's dive in!

1. Moments: The "Snapshots" of a Distribution

In physics, a "moment" relates to the position and rotation of an object. In probability, a moment is a quantitative measure of the shape of a set of points. We categorize them into two types: Raw Moments and Central Moments.

The k-th Raw Moment

The k-th raw moment is simply the expected value of the random variable \( X \) raised to the power of \( k \). We usually denote the mean as \( \mu \).

Formula: \( E[X^k] \)
• If \( k = 1 \), it’s the 1st raw moment: \( E[X^1] \), which is just the Mean (\( \mu \)).
• If \( k = 2 \), it’s the 2nd raw moment: \( E[X^2] \).

The k-th Central Moment

A central moment is calculated relative to the mean (\( \mu \)). It tells us how the data "swings" around its center.

Formula: \( E[(X - \mu)^k] \)
• The 1st central moment is always 0 because the average deviation from the average is always zero! \( E[X - \mu] = E[X] - \mu = 0 \).
• The 2nd central moment is the most famous one: Variance.

Quick Review Box:
Raw Moment: Centered around zero (\( E[X^k] \))
Central Moment: Centered around the mean (\( E[(X - \mu)^k] \))

2. Variance and Standard Deviation

Variance measures the "spread" or "volatility" of a random variable. In the insurance world, high variance means high risk because the actual claims could be much higher (or lower) than expected.

The "Must-Know" Formula:
\( Var(X) = E[X^2] - (E[X])^2 \)

Memory Trick: To find the variance, take "The Mean of the Squares" minus "The Square of the Mean."

Standard Deviation (\( \sigma \)):
Variance is measured in "squared units" (like dollars squared), which is hard to visualize. We take the square root to get back to the original units.
\( \sigma = \sqrt{Var(X)} \)

Common Mistake to Avoid:
Students often confuse \( E[X^2] \) with \( (E[X])^2 \). They are very different! Always calculate \( E[X^2] \) by integrating \( x^2 \cdot f(x) \) first, then subtract the mean squared.

Key Takeaway: Variance is the 2nd central moment. It tells us how far the "average" outcome is from the mean.

The Mode is the value of \( x \) that is most likely to occur. On a graph of a probability density function (pdf), the mode is the "peak" of the mountain.

How to find it:
1. For Discrete variables: Look for the \( x \) with the highest probability \( P(X=x) \).
2. For Continuous variables: Find the value of \( x \) that maximizes the density function \( f(x) \). This usually involves taking the derivative \( f'(x) \), setting it to 0, and solving for \( x \).

Did you know? A distribution can have more than one mode (bimodal) if there are two peaks of the same height!

4. Median and Percentiles

Sometimes the "average" (mean) is misleading if there are extreme outliers (like a billionaire walking into a local bar). In those cases, we use Percentiles.

The Median (The 50th Percentile)

The Median is the middle value. Half of the data is below it, and half is above it. It is often denoted as \( m \) or \( \pi_{0.5} \).

The Calculation:
Solve for \( m \) in this equation:
\( F(m) = P(X \le m) = 0.5 \)

General Percentiles

The p-th percentile (denoted \( \pi_p \)) is the value such that the probability of being less than or equal to that value is \( p \).

Step-by-Step Process for Continuous Variables:
1. Find the Cumulative Distribution Function (CDF), \( F(x) \).
2. Set \( F(\pi_p) = p \).
3. Solve for \( \pi_p \).

Example: If you want the 75th percentile (the 3rd quartile), you solve \( F(x) = 0.75 \).

Analogy: Imagine a long line of people ranked by height. If you are the 90th percentile, you are taller than 90% of the people in that line.

Quick Review Box:
Mean: The balance point (Arithmetic average).
Median: The 50/50 split point.
Mode: The most frequent point (The peak).

5. Bringing it All Together: Comparing Mean, Median, and Mode

The relationship between these three tells you about the Skewness (the "lean") of your distribution:

• Symmetric: Mean = Median = Mode (like a perfect Bell Curve).
• Right-Skewed (Positive Skew): Mean > Median > Mode. This has a "long tail" on the right side. (Common in insurance claims!)
• Left-Skewed (Negative Skew): Mean < Median < Mode. This has a "long tail" on the left side.

Key Takeaway for Exam P: The SOA loves to ask you to calculate these for specific distributions like the Exponential or Uniform. Always start by identifying if the variable is discrete or continuous, then choose your tool (summation for discrete, integration for continuous).

Final Encouragement

This chapter is all about summarizing data. If the formulas for moments look scary, just remember they are just different ways of "weighting" the values of \( X \). Practice calculating the 2nd raw moment \( E[X^2] \) for different functions—once you master that, Variance and Standard Deviation will become second nature. You've got this!