Skip to content
Tutorials Center
My learning

Topics

Lesson 26 of 37 · Probability Distributions

Z-Scores and the Standard Normal Distribution

Standardize values with z = (x - mean) / sd, find probabilities and percentiles for normal data, and approximate binomial probabilities with a continuity correction.

A score of 80 means very different things on a test averaging 50 and on a test averaging 75. To compare values from different scales, express each one as the number of standard deviations from its mean. That number is the z-score, and it lets one table (or one function) answer every normal-distribution question.

This lesson continues from the normal distribution.

Standardizing

Text
z = (x - μ) / σ

If X ~ N(μ, σ²), then Z = (X - μ)/σ follows the standard normal distribution N(0, 1): mean 0, standard deviation 1. A z of +1.5 means "1.5 standard deviations above the mean"; negative z is below the mean.

Example

An exam has mean 62 and standard deviation 12. A student scores 80:

Text
z = (80 - 62) / 12 = 1.5

From the standard normal, Φ(1.5) ≈ 0.9332, so about 93.3% of students scored below 80, and P(X > 80) ≈ 0.0668.

Python A z-score and probability calculator

Press Run (or Ctrl + Enter) to see the result here.

Program input · what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Comparing across scales

Maya scores 84 on a math test (mean 70, sd 8). Leo scores 130 on a reading test (mean 100, sd 20). Who did better relative to their group?

  • Maya: z = 14/8 = 1.75.
  • Leo: z = 30/20 = 1.5.

Maya's performance is further above her group's average.

Finding percentiles: working backwards

Sometimes the probability is given and you need x. Find the z with the required cumulative area (the inverse of Φ), then convert back: x = μ + zσ. For the exam above, the 90th percentile has z ≈ 1.2816, so x ≈ 62 + 12 × 1.2816 ≈ 77.4.

Python's standard library can do this:

Python Percentiles with the inverse normal CDF

Press Run (or Ctrl + Enter) to see the result here.

Program input · what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Notice that the 97.5th percentile of the standard normal is about 1.96. That is where the number 1.96 in confidence intervals comes from.

Normal approximation to the binomial

When n is large, B(n, p) looks like a normal curve with μ = np and σ = √(np(1-p)). The approximation works well when both np and n(1-p) are at least about 10 (a common guideline).

Because the binomial is discrete and the normal continuous, apply a continuity correction: count each integer k as the interval from k - 0.5 to k + 0.5. So P(X ≤ 55) becomes P(Y < 55.5).

Example: X ~ B(100, 0.5), μ = 50, σ = 5:

Text
z = (55.5 - 50) / 5 = 1.1,  Φ(1.1) ≈ 0.864
Python How good is the normal approximation?

Press Run (or Ctrl + Enter) to see the result here.

Program input · what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

When using a z-table, round z to two decimals. With software, keep full precision and round only the final answer.

Check your understanding

Quick check · 4 questions

of 4 answered You scored / 4 · perfect!

  1. 1. A value x = 90 comes from a distribution with mean 75 and standard deviation 5. What is its z-score?

    z = (90 - 75) / 5 = 3.
  2. 2. A z-score of -2 means the value is...

    z measures distance from the mean in standard deviations, and negative means below.
  3. 3. The 97.5th percentile of the standard normal distribution is approximately...

    About 95% lies between -1.96 and 1.96, leaving 2.5% in each tail.
  4. 4. Why use a continuity correction when approximating a binomial with a normal curve?

    Each integer k occupies the interval from k - 0.5 to k + 0.5, so P(X ≤ k) uses the boundary k + 0.5.

Summary

  • z = (x - μ)/σ counts standard deviations from the mean.
  • Standardizing turns any normal question into one about N(0, 1).
  • Percentiles reverse the process: find z, then x = μ + zσ.
  • Large binomials are approximately normal, with a continuity correction of 0.5.

Next: Scatter Plots and Correlation