01 / 20

Introducing Statistics
Why Be Normal?

Explore the normal distribution, learn to describe data with \(\mu\) and \(\sigma\), and discover why the bell curve appears everywhere in the real world.

AP Statistics

What is Statistics?

Statistics is the science of collecting, organizing, analyzing, and interpreting data to make informed decisions. Rather than examining every individual, we use statistics to draw conclusions about an entire population from a smaller sample.

Population vs. Sample

Population
The entire group you want to study. Parameters use Greek letters: mean \(\mu\), standard deviation \(\sigma\).
Sample
A subset drawn from the population. Statistics use Latin letters: mean \(\bar{x}\), standard deviation \(s\).

Describing Data: Center and Spread

Mean \(\mu\)
The arithmetic average: sum of all values divided by the count. Sensitive to outliers.
Median
The middle value when data is sorted. Resistant to outliers and skew.
Std Dev \(\sigma\)
Measures how far values typically fall from the mean. Larger \(\sigma\) means more spread.

The Normal Distribution

The 68-95-99.7 Rule

1
68% of data falls within \(\mu \pm 1\sigma\)
2
95% of data falls within \(\mu \pm 2\sigma\)
3
99.7% of data falls within \(\mu \pm 3\sigma\)

\(z\)-Scores and the Standard Normal

A \(z\)-score tells you how many standard deviations a value is from the mean. Converting to \(z\)-scores lets us compare values from different distributions on the same scale.

Normal vs. Skewed Distributions

Symmetric (Normal)
\(\text{Mean} \approx \text{Median}\). Tails are mirror images. The 68-95-99.7 rule applies.
Right-Skewed
Long tail stretches right. \(\text{Mean} > \text{Median}\). Outliers pull the mean up.
Left-Skewed
Long tail stretches left. \(\text{Mean} < \text{Median}\). Outliers pull the mean down.

When Does Normal Apply?

Not all data is normal. The normal model works well when the data is unimodal (one peak), symmetric, and free of strong outliers. Always check with a histogram or normal probability plot before assuming normality.

1
Heights of adult humans: approximately normal
2
Test scores on a large standardized exam: often approximately normal
3
Income data: typically right-skewed, not normal
4
Measurement errors in lab experiments: approximately normal

Applying the Normal Model

1
Identify the variable and confirm the distribution is approximately normal (histogram, normal plot).
2
State the parameters: write \(X \sim N(\mu, \sigma)\) with the given or estimated values.
3
Standardize: convert your value to a \(z\)-score using \(z = \frac{x - \mu}{\sigma}\).
4
Use the table or calculator: find the area (probability) from the standard normal table or normalcdf on your calculator.
5
Interpret in context: state the probability as a percentage and relate it back to the original question.

Example 1: Finding a Probability

Adult women's heights are approximately normal with \(\mu = 64\) inches and \(\sigma = 2.5\) inches. What proportion of women are taller than 69 inches?

1
Set up: \(X \sim N(64, 2.5)\). We want \(P(X > 69)\).
2
Standardize: \(z = \frac{69 - 64}{2.5} = \frac{5}{2.5} = 2.0\)
3
Look up: \(P(Z > 2.0) = 1 - 0.9772 = 0.0228\)

Example 2: Working Backwards

On a standardized test, scores follow \(N(500, 100)\). What score do you need to be in the top 10%?

1
Find the \(z\)-score: We need \(P(Z > z) = 0.10\), so \(P(Z \leq z) = 0.90\). From the table, \(z \approx 1.28\).
2
Un-standardize: \(x = \mu + z\sigma = 500 + 1.28(100) = 628\).

Example 3: Between Two Values

Battery life is \(N(300, 20)\) hours. What proportion of batteries last between 260 and 340 hours?

1
Standardize both: \(z_1 = \frac{260 - 300}{20} = -2.0\), \(z_2 = \frac{340 - 300}{20} = 2.0\).
2
Find area: \(P(-2 \leq Z \leq 2) = 0.9772 - 0.0228 = 0.9544\).

Common Mistakes

✕ Assuming Normality
Applying the normal model to clearly skewed data. Always check shape first with a graph.
✕ Wrong Tail Direction
Forgetting to subtract from 1. If you need \(P(X > x)\), compute \(1 - P(Z \leq z)\).
✕ Mixing Up \(\sigma\) and \(\sigma^2\)
Using variance \(\sigma^2\) where standard deviation \(\sigma\) is needed, or vice versa.

Summary Reference

Concept Key Formula / Value
Mean Center of distribution \(\mu = \frac{\sum x_i}{N}\)
Std Dev Spread of distribution \(\sigma = \sqrt{\frac{\sum(x_i - \mu)^2}{N}}\)
\(z\)-score Standardized value \(z = \frac{x - \mu}{\sigma}\)
68-95-99.7 Empirical Rule \(\pm 1\sigma, \pm 2\sigma, \pm 3\sigma\)
Normal PDF Bell curve formula \(f(x)=\frac{1}{\sigma\sqrt{2\pi}}e^{-\frac{(x-\mu)^2}{2\sigma^2}}\)

Question 1 of 5

A dataset follows \(N(50, 5)\). What is the \(z\)-score for a value of 62.5?

Question 2 of 5

According to the 68-95-99.7 rule, what percentage of data in a normal distribution falls within one standard deviation of the mean?

Question 3 of 5

If IQ scores follow \(N(100, 15)\), what is the probability that a randomly selected person has an IQ greater than 130?

Question 4 of 5

In a right-skewed distribution, which relationship is typically true?

Question 5 of 5

On a test with \(\mu = 72\) and \(\sigma = 8\), a student's \(z\)-score is \(-1.5\). What is the student's raw score?