Testing hypotheses

Home > Subject areas > Studying statistics > Testing hypotheses

Testing hypotheses is one of the most widespread uses of statistical methods. Strictly, we should talk of null hypothesis testing, because the hypotheses tested are always null ones – what this means should become clearer when you look at the examples. (The tests are sometimes referred to as significance tests.)

The basic idea of statistical hypothesis testing is very simple. However, the details of how to carry out a test – the mathematical formulae, statistical tables, and so on – tend to be complicated. And you will find that each different situation requires a different test, with different formulae and different tables. Furthermore, the interpretation of the results is often confusing, and mistakes and misinterpretations are common.

• Hypothesis test 1: an experiment on telepathy
• Hypothesis test 2: the shuffle test
• Basic format of hypothesis test
• The interpretation of hypothesis tests
• Exercises
• Useful resources

This content has been written by Michael Wood author of Making Sense of Statistics