Introduction to the Normal Distribution

In your study of Statistics, you've looked at how data can be spread out in many ways. However, one specific pattern appears so often in nature, science, and even exam results that we call it the Normal distribution. You might have heard it called the "Bell Curve" because of its distinctive shape.

In this chapter, we will explore why this distribution is so special, how to recognize it, and how to use its predictable properties to solve probability problems. Don't worry if the symbols look a bit strange at first—once you see the pattern, it's very logical!

What makes a Distribution "Normal"?

A Normal distribution isn't just any curve; it has very specific characteristics. For a set of data to be modeled by a Normal distribution, it must be:

  • Continuous: It deals with data that can take any value (like height, weight, or time), not just whole numbers.
  • Symmetrical: If you folded the curve in half at the center, both sides would match perfectly.
  • Bell-shaped: It has one high peak in the middle and "tails" that go off to the left and right.
  • Centered on the Mean: In a perfect Normal distribution, the Mean, Median, and Mode are all exactly the same value and are located right in the center.

The Notation: \(N(\mu, \sigma^2)\)

Statisticians use a shorthand to describe a Normal distribution. You will see it written like this: \(X \sim N(\mu, \sigma^2)\).

Let’s break that down:

  • \(X\): This is just the variable we are measuring (e.g., the weight of an apple).
  • \(\sim\): This symbol means "follows the distribution of."
  • \(N\): This tells us the distribution is Normal.
  • \(\mu\) (mu): This Greek letter represents the population mean. It tells you where the center of the bell is.
  • \(\sigma^2\) (sigma squared): This represents the variance. It tells you how spread out the bell is.

Important Tip: Remember that \(\sigma\) (without the square) is the standard deviation. If a question gives you \(N(50, 16)\), the mean is \(50\) and the standard deviation is \(\sqrt{16} = 4\).

Key Takeaway

The Normal distribution is defined entirely by its mean (\(\mu\)) and its standard deviation (\(\sigma\)). If you know these two numbers, you know everything about the shape and position of the curve!

When is a Normal Model Suitable?

You might be asked if a Normal distribution is a "suitable model" for a set of data. It is suitable if:

1. The data is continuous.
2. The distribution is symmetrical (not skewed).
3. Most of the data is clustered around the mean, with fewer values as you move further away.

Example: The heights of adult women in the UK usually follow a Normal distribution. Most women are near the average height, with very few being extremely short or extremely tall.

The "68% - 95% - 99.7%" Rule

One of the most powerful things about the Normal distribution is how predictable it is. No matter what the mean or standard deviation is, the area under the curve (which represents probability) always follows these rules:

  • 68% of the data lies within one standard deviation of the mean: \((\mu \pm 1\sigma)\).
  • 95% of the data lies within two standard deviations of the mean: \((\mu \pm 2\sigma)\).
  • 99.7% (almost all) of the data lies within three standard deviations of the mean: \((\mu \pm 3\sigma)\).

Example Problem:
The weights of bags of sugar are normally distributed with a mean \(\mu = 1000g\) and a standard deviation \(\sigma = 5g\).
Question: What percentage of bags weigh between \(990g\) and \(1010g\)?
Solution:
1. Calculate how many standard deviations \(990g\) and \(1010g\) are from the mean.
2. \(1000 - 990 = 10\) (which is \(2 \times 5g\), so \(2\sigma\)).
3. \(1010 - 1000 = 10\) (which is \(2 \times 5g\), so \(2\sigma\)).
4. According to our rule, 95% of the data falls within \(2\) standard deviations. So, 95% of the bags weigh between \(990g\) and \(1010g\).

Standardising Data

While you aren't expected to use complex statistical tables in this exam, you are expected to understand standardised scores. This allows us to compare two different Normal distributions by seeing how many standard deviations a value is away from the mean.

The formula for a standardised score (often called a z-score) is:
\(Standardised Score = \frac{x - \mu}{\sigma}\)

Where:
\(x\) is the value you are looking at.
\(\mu\) is the mean.
\(\sigma\) is the standard deviation.

Quick Review: We also use these scores in the Quality Assurance chapter to set warning and action lines on control charts!

Common Pitfalls to Avoid

  • Confusing \(\sigma\) and \(\sigma^2\): In the notation \(N(\mu, \sigma^2)\), the second number is the variance. Always square root it to get the standard deviation before applying the 68-95-99.7 rule.
  • Assuming symmetry when it's skewed: If a data set has a "tail" longer on one side (skewed), the Normal distribution is not a suitable model.
  • Mixing up Binomial and Normal: The Binomial distribution is for discrete data (things you count), while the Normal distribution is for continuous data (things you measure).

Quick Review Summary

1. Notation: \(N(\mu, \sigma^2)\) where \(\mu\) is the mean and \(\sigma^2\) is the variance.
2. Shape: Symmetrical bell curve where Mean = Median = Mode.
3. The Rules:
- \(\mu \pm 1\sigma \approx 68\%\)
- \(\mu \pm 2\sigma \approx 95\%\)
- \(\mu \pm 3\sigma \approx 99.7\%\)
4. Suitability: Use for continuous, symmetrical data that clusters in the middle.