Course contentsShow
← Mathematics
Lesson 3026 of 3,12159. Statistics — Descriptive and InferentialFree lesson

Two-Tailed z-Tests

Conducting hypothesis tests for μ ≠ μ₀ and interpreting results with two-sided rejection regions.

Two-Tailed z-Tests

What you'll learn: How to test whether a population mean differs from a hypothesized value in either direction using a two-sided rejection region.

The Core Idea

In the previous lesson, you learned the one-sample z-test where the alternative hypothesis pointed in one direction (μ > μ₀ or μ < μ₀). A two-tailed z-test is used when you want to detect a difference in either direction—that is, when your alternative hypothesis is H₁: μ ≠ μ₀.

Think of it like a security alarm that triggers whether someone is too tall or too short to match a description. You're not just watching for one extreme; you're watching both.

How It Works

The setup is familiar:

  • Null hypothesis: H₀: μ = μ₀
  • Alternative hypothesis: H₁: μ ≠ μ₀
  • Test statistic: z = (x̄ - μ₀)/(σ/√n)

The key difference is the rejection region. Since you're testing for difference in either direction, you split your significance level α between both tails of the standard normal distribution. For example, at α = 0.05, you reject H₀ if:

  • z < -1.96 (left tail, 2.5% of probability), or
  • z > +1.96 (right tail, 2.5% of probability)

For the p-value approach, calculate P(|Z| ≥ |z_observed|) = 2·P(Z ≥ |z_observed|). You're doubling the tail probability because evidence could appear on either side.

When to Use Two-Tailed Tests

Use a two-tailed test when you genuinely care about detecting change in either direction, with no prior reason to focus on one side.

Key Takeaway: A two-tailed z-test checks for μ ≠ μ₀ by splitting the rejection region between both tails of the normal distribution, requiring stronger evidence in any one direction compared to a one-tailed test at the same α level.