Normal distribution

Introduction: The Shape of the World

Welcome to one of the most famous topics in all of Statistics! Have you ever noticed that most people are "average" height, and very few are exceptionally tall or short? Or that most apples in a bag weigh about the same, with only a few being tiny or massive? This "natural" pattern is what we call the Normal Distribution.

In this chapter, we will learn how to model these patterns using math. For your OCR MEI H640 exam, this is a vital tool because it allows us to calculate probabilities for things that happen in real life. Don't worry if this seems tricky at first—once you see the "Bell Curve" shape, it all starts to click!

1. What is the Normal Distribution?

The Normal Distribution is a continuous probability distribution. This means it deals with data that can take any value (like time, weight, or height), rather than "counting" data (like the number of goals scored).

Key Characteristics

• Bell-Shaped: It is perfectly symmetrical around the middle.
• The Mean (\(\mu\)): The peak of the curve is exactly at the mean. This is also the median and the mode!
• Points of Inflection: The curve changes from concave downward to concave upward exactly one standard deviation (\(\sigma\)) away from the mean (at \(\mu \pm \sigma\)).
• Total Area: The total area under the curve is exactly 1 (representing a total probability of 100%).

The Notation

We write it like this: \(X \sim N(\mu, \sigma^2)\)
Where:
- \(X\) is our random variable.
- \(\mu\) is the mean.
- \(\sigma^2\) is the variance (remember: \(\sigma\) is the standard deviation).

Quick Review Box:
If you see \(X \sim N(50, 16)\), the mean is 50 and the standard deviation is \(\sqrt{16} = 4\). Common Mistake: Students often forget to square root the second number to get the standard deviation!

2. The "Bell Curve" Shape and Symmetry

Because the curve is symmetrical, we know that 50% of the data is below the mean and 50% is above the mean. This makes calculations much easier!

Did you know?

In a Normal Distribution:
- About 68% of the data lies within 1 standard deviation of the mean (\(\mu \pm \sigma\)).
- About 95% lies within 2 standard deviations (\(\mu \pm 2\sigma\)).
- About 99.7% lies within 3 standard deviations (\(\mu \pm 3\sigma\)).

3. Calculating Probabilities

In your exam, you won't have to do the calculus behind this curve. Instead, you will use your statistical calculator.

How to use your calculator:

1. Find the Distribution menu.
2. Select Normal CD (Cumulative Distribution).
3. Enter your Lower bound, Upper bound, \(\sigma\) (standard deviation), and \(\mu\) (mean).
4. Example: To find \(P(X < 60)\) when \(X \sim N(50, 16)\), your Lower would be a very small number (like -9999) and your Upper would be 60.

Analogy: Think of the probability as the amount of paint needed to cover the area under the curve between two points. The more area you cover, the higher the probability!

The Continuity Correction

The syllabus mentions that we sometimes use the Normal distribution to model discrete data (like the number of people who pass a test). Because the Normal distribution is continuous, we have to make a small adjustment.

If you want to find the probability of "at least 10" in discrete data, you would look at the area from 9.5 upwards in the Normal model. Imagine each whole number occupies a "box" that stretches 0.5 units on either side.

Key Takeaway: Use the continuity correction only when you are using a continuous curve (Normal) to estimate discrete counts (like Binomial data).

4. Standardizing: The Z-score

Sometimes we want to find unknown values of \(\mu\) or \(\sigma\), or compare two distributions. To do this, we "Standardize" into a Standard Normal Distribution, which we call \(Z\).

\(Z \sim N(0, 1)\) (Mean is 0, Variance is 1).

The Formula:

\(Z = \frac{X - \mu}{\sigma}\)

This tells you how many standard deviations a value is away from the mean.

5. Linear Transformations of Normal Variables

If \(X\) follows a Normal distribution and we apply a linear transformation \(Y = aX + b\), then \(Y\) also follows a Normal distribution:

If \(X \sim N(\mu, \sigma^2)\), then \(Y = aX + b \sim N(a\mu + b, a^2\sigma^2)\).

Notice that the mean changes by the linear formula \(a\mu + b\), while the variance is multiplied by \(a^2\) (the constant addition \(b\) does not affect the spread).

6. Hypothesis Testing for the Mean

This is a core component of the MEI syllabus. We want to test whether the mean of a sample provides evidence that the population mean has changed.

The Distribution of the Sample Mean

If we take a random sample of size \(n\) from \(X \sim N(\mu, \sigma^2)\), the sample mean (\(\bar{X}\)) follows its own Normal distribution:
\(\bar{X} \sim N\left(\mu, \frac{\sigma^2}{n}\right)\)

Crucial Point: The standard error of the sample mean is \(\frac{\sigma}{\sqrt{n}}\). As sample size \(n\) increases, the spread of the sample mean narrows.

Step-by-Step Hypothesis Test:

1. State Hypotheses: Define the population mean \(\mu\). State \(H_0: \mu = \mu_0\) and \(H_1: \mu > \mu_0\), \(\mu < \mu_0\), or \(\mu \neq \mu_0\).
2. State Distribution under \(H_0\): Under the null hypothesis, \(\bar{X} \sim N\left(\mu_0, \frac{\sigma^2}{n}\right)\).
3. Choose Method (p-value or Critical Region):
- p-value method: Calculate \(P(\bar{X} \ge \bar{x})\) or \(P(\bar{X} \le \bar{x})\). If the p-value is less than the significance level \(\alpha\) (or \(\frac{\alpha}{2}\) for a 2-tailed test), reject \(H_0\).
- Critical Region method: Find the cutoff value(s) of \(\bar{X}\) corresponding to the tail probability \(\alpha\). If your observed sample mean falls in the critical region, reject \(H_0\).
4. Conclusion in Context: Always give a non-assertive conclusion referring back to the context of the question (e.g., "There is sufficient evidence at the 5% significance level to suggest that the mean weight of apples has decreased.").

Common Mistake: Forgetting to divide the variance by \(n\) when conducting a hypothesis test for the sample mean. Always check your sample size!

Summary: What you must remember

• The Model: \(X \sim N(\mu, \sigma^2)\). The curve is bell-shaped and symmetrical around \(\mu\).
• Standardizing: \(Z = \frac{X - \mu}{\sigma}\) converts any normal distribution to \(Z \sim N(0, 1)\).
• Linear Combinations: If \(X \sim N(\mu, \sigma^2)\), then \(aX + b \sim N(a\mu + b, a^2\sigma^2)\).
• Sample Means: For a sample of size \(n\), \(\bar{X} \sim N\left(\mu, \frac{\sigma^2}{n}\right)\).
• Hypothesis Tests: Use either the \(p\)-value or critical value method with standard error \(\frac{\sigma}{\sqrt{n}}\), and write your conclusion in context.

Keep practicing those calculator steps! Once you are comfortable with the functions, the Normal Distribution becomes one of the most predictable and high-scoring sections of your Statistics paper.