Intro

At the root of many approaches to experimental design in science is a hypothesis test. Statistical hypothesis testing is used to make decisions about an experiment in the presence of variability.

The key terms we address in this section are:

• Null hypothesis
• Alternative hypotheses
• Type I error
• Type II error

The Null and Alternative Hypotheses

The null hypothesis asserts that there is no difference or effect arising from experimental conditions or settings in a scientific investigation.

The alternative hypothesis asserts that there is a scientifically relevant difference or effect arising from the experimental conditions.

For example, a researcher wants to know if coffee positively affects test scores. The null hypothesis is that coffee does not affect scores. The alternative is that coffee does have an effect.