Hypothesis Testing
The heart of inferential statistics. We use data to decide whether to accept or reject a claim.
Null vs Alternative Hypothesis
- Null Hypothesis ($H_0$): The status quo. "No difference", "No effect". (e.g., $\mu = 50$).
- Alternative Hypothesis ($H_1$ or $H_a$): The claim we want to test. (e.g., $\mu \ne 50$ or $\mu > 50$).
The 5 Steps
- State $H_0$ and $H_1$.
- Select Level of Significance ($\alpha$, usually 0.05).
- Calculate the Test Statistic (Z or t).
- Determine the Critical Value or P-Value.
- Make a Decision (Reject or Fail to Reject $H_0$).
Z-Test for Mean ($\sigma$ known)
$$ Z = \frac{\bar{x} - \mu}{\sigma / \sqrt{n}} $$
t-Test for Mean ($\sigma$ unknown)
Used when $\sigma$ is unknown. We use $s$ (sample std dev) instead.
$$ t = \frac{\bar{x} - \mu}{s / \sqrt{n}} $$
Advanced Topics
Dive deeper into the nuances of testing:
Test Yourself
Q1: If we Reject $H_0$ when it is actually true, what error have we made?