01 / 21

The Normal Distribution
Revisited

Deepen your understanding of the bell curve, master z-scores, and apply the Empirical Rule to solve real-world probability problems.

AP Statistics

What Is the Normal Distribution?

The normal distribution is a continuous probability distribution that is symmetric about its mean, producing the classic bell-shaped curve. It is defined by two parameters: the mean \(\mu\) and the standard deviation \(\sigma\).

\(\mu\)
\(\mu - \sigma\)
\(\mu + \sigma\)
Peak at \(\mu\)

The Two Parameters: \(\mu\) and \(\sigma\)

\(\mu\) : Mean
Centers the distribution. Shifting \(\mu\) slides the entire bell curve left or right along the number line without changing its shape.
\(\sigma\) : Standard Deviation
Controls the spread. A smaller \(\sigma\) produces a taller, narrower curve; a larger \(\sigma\) produces a shorter, wider curve.

Why Does It Matter?

The normal distribution appears everywhere: test scores, heights, measurement errors, and countless natural phenomena. It is the foundation of inferential statistics because of the Central Limit Theorem.

1
Modeling real data: Many real-world variables are approximately normal.
2
Probability calculations: We can find exact probabilities for any interval using the standard normal table.
3
Inference: Confidence intervals and hypothesis tests rely on the normal (or \(t\)) distribution.

The Empirical Rule: 68-95-99.7

68% 95% 99.7%
\(\mu - \sigma\)
\(\mu + \sigma\)
\(\mu - 2\sigma\)
\(\mu + 2\sigma\)
\(\mu\)

The Standard Normal and Z-Scores

The standard normal distribution has \(\mu = 0\) and \(\sigma = 1\). We convert any normal variable \(X\) to a z-score to use the standard normal table:

\(z > 0\)
Value is above the mean
\(z = 0\)
Value equals the mean exactly
\(z < 0\)
Value is below the mean

Normal vs Standard Normal

Normal \(N(\mu, \sigma)\) Standard Normal \(N(0, 1)\)
Mean Any value \(\mu\) \(0\)
Std Dev Any value \(\sigma > 0\) \(1\)
Variable \(X\) \(Z\)
Table Lookup Must convert to \(z\) first Use directly
Use Case Model raw data Probability calculations

How to Find Normal Probabilities

1
State the problem: Identify \(\mu\), \(\sigma\), and the value(s) of interest.
2
Draw a picture: Sketch the normal curve and shade the region of interest.
3
Standardize: Compute \(z = \frac{X - \mu}{\sigma}\) for each boundary value.
4
Use the table or calculator: Look up \(P(Z \leq z)\) to find the cumulative area to the left.
5
Adjust if needed: For right-tail: \(1 - P(Z \leq z)\). For between: subtract the two cumulative areas.

Example 1: Left-Tail Probability

SAT scores are normally distributed with \(\mu = 500\) and \(\sigma = 100\). What is the probability that a randomly chosen student scores below 620?

620
\(z=1.20\)
500 0.8849

Example 2: Right-Tail Probability

The weight of apples at a farm is normally distributed with \(\mu = 150\) g and \(\sigma = 20\) g. What proportion weigh more than 180 g?

180 g
\(z=1.50\)
150 g 0.0668

Example 3: Between Two Values

Heights of adult women are \(N(64, 2.5)\) inches. Find \(P(60 \leq X \leq 67)\).

60" 67" 64" 0.8301

Working Backwards: Inverse Normal

Sometimes we know the probability and need to find the value. This is the inverse normal problem: given an area, find the corresponding \(X\).

1
Identify the area: Determine the cumulative area to the left (or convert if given a right-tail or percentile).
2
Find \(z^*\): Use the z-table in reverse (or invNorm on your calculator) to find the z-score matching that area.
3
Unstandardize: Solve \(X = \mu + z^* \cdot \sigma\) to get the original-scale value.

Assessing Normality

Before using normal calculations, we should check whether the data are approximately normal. There are several tools for this.

Histogram
Should appear roughly symmetric and bell-shaped, without strong skewness or multiple peaks.
Normal Probability Plot
Points should fall approximately along a straight line. Systematic curvature indicates non-normality.
Empirical Rule Check
Verify that roughly 68% of data falls within 1 SD, 95% within 2 SD, and 99.7% within 3 SD of the mean.

A filling line that deposits one statistic for a whole tray

A filling line loads bottles into trays while one statistic is recorded for each tray.A tray contains individual observations, but its average contributes one value to a distribution of sample means. Repeating the sampling process reveals how that statistic varies.

Build the distribution of tray means

Twenty-six possible sample means appear as discrete probability bars with a normal approximation drawn for comparison.For 25 independent fills equally likely to be 40 or 60 mL, K is binomial with n = 25 and p = 0.5, and the tray mean is 40 + 0.8K. The exact distribution is discrete, centered at 50 mL with standard deviation 2 mL. A normal curve approximates its shape.

Reading a Normal Probability Plot

Approximately Normal Right-Skewed Expected z-score Expected z-score Points on line = Normal Curved pattern = Not Normal

Common Mistakes to Avoid

✕ Forgetting to Standardize
Using raw \(X\) values directly in the z-table instead of computing \(z = \frac{X - \mu}{\sigma}\) first.
✕ Wrong Tail
The z-table gives \(P(Z \leq z)\). For "greater than" questions, you must subtract from 1.
✕ Confusing \(\sigma\) with \(\sigma^2\)
Some textbooks write \(N(\mu, \sigma^2)\). Make sure you use the standard deviation, not the variance, in the z-formula.

Summary of Key Formulas

Concept Formula / Rule When to Use
Z-score \(z = \frac{X - \mu}{\sigma}\) Convert to standard normal
Unstandardize \(X = \mu + z\sigma\) Inverse normal problems
Empirical Rule 68 / 95 / 99.7 Quick estimates within 1, 2, 3 SD
Left-tail \(P\) \(P(Z \leq z)\) from table "Less than" questions
Right-tail \(P\) \(1 - P(Z \leq z)\) "Greater than" questions
Between \(P(z_1 \leq Z \leq z_2)\) Subtract two cumulative areas

Question 1 of 5

A distribution is \(N(80, 5)\). Using the Empirical Rule, approximately what percentage of values fall between 70 and 90?

Question 2 of 5

If \(X \sim N(200, 25)\) and \(X = 250\), what is the z-score?

Question 3 of 5

\(P(Z \leq 1.00) = 0.8413\). What is \(P(Z > 1.00)\)?

Question 4 of 5

A normal probability plot of a dataset shows points that curve sharply upward at the right end. This suggests the data are:

Question 5 of 5

Scores on a test are \(N(72, 8)\). A student wants to be in the top 5%. What is the minimum score needed? (Use \(z^* = 1.645\) for the 95th percentile.)