7 Hypothesis Testing
8 What is a hypothesis?
A statistical framework to decide whether the evidence from a sample is strong enough to make conclusions about the population.
- A way to answer research questions with data.
- Starts with a null hypothesis (H₀): no effect, no difference.
- Competes against an alternative hypothesis (H₁): there is an effect or difference.
- Uses probability to control for random chance.
8.1 Why do we need this?
Studying descriptive statistics gives a view on the data at hand. But to make conclusions about the data, connect them to the real world patterns we need inference statistics and the major contribution comes from hypothesis testing.
8.1.1 Concepts to know to proceed with Hypothesis testing
- Null Hypothesis
- Alternate Hypothesis
- p-value
- False Discovery Rate
- Effect Size
- Error rates
- Power
- Multiple Comparisons