Setting Up Hypotheses for One-Sample t-Test
You'll learn: How to formulate the null and alternative hypotheses when testing whether a population mean equals a specific value using a one-sample t-test.
The Core Idea
When performing a one-sample t-test, you're asking: "Does my population mean differ from a specific value I care about?" Before you calculate anything, you must clearly state two competing claims:
- Null hypothesis (H₀): The population mean equals your reference value
- Alternative hypothesis (H₁): The population mean differs from that value (or is greater/less than it)
Think of it like a courtroom: H₀ is the "presumption of innocence" (nothing unusual is happening), and H₁ is the claim you're investigating.
Writing the Hypotheses
For a one-sample t-test comparing a population mean (μ) to a specific value (μ₀):
Two-tailed test (testing for any difference):
- H₀: μ = μ₀
- H₁: μ ≠ μ₀
One-tailed test (testing for a specific direction):
- H₀: μ = μ₀
- H₁: μ > μ₀ (or μ < μ₀)
Example in Context
Suppose a cereal box claims to contain 500g. You collect a sample and want to test if the actual mean weight differs from 500g:
- H₀: μ = 500 (the company's claim is accurate)
- H₁: μ ≠ 500 (the actual mean differs from the claim)
Notice how H₀ always contains the "equals" sign. You're testing against a specific value, and the burden of proof lies on showing the data are inconsistent with that value.
Key Takeaway: In a one-sample t-test, H₀ states that the population mean equals your reference value (μ = μ₀), while H₁ states it differs (≠), is greater (>), or is less (<) than that value, depending on your research question.