Erreurs dans les tests d'hypothèse

A Type I error occurs when you reject a null hypothesis that is actually true, leading to an incorrect and misleading conclusion. The probability of committing a Type I error is represented by the Greek letter α (alpha), commonly set at 0.05 or 0.01. This acceptable probability value is also known as the significance level.

A Type II error occurs when you fail to reject a null hypothesis that is actually false. Unlike Type I error, which rejects a true null hypothesis, Type II error represents a false negative outcome. The probability of Type II error is denoted by β (beta) and is calculated from the power of the hypothesis test.

Correct decisions occur in two scenarios: rejecting the null hypothesis when it is actually false, or failing to reject it when it is actually true. The probability of correctly rejecting a false null hypothesis is called the power of the test, which equals 1 minus β. Ideally, power should be as close to one as possible.

The significance level, represented by α (alpha), is the acceptable probability of committing a Type I error. Commonly set at 0.05 or 0.01, it defines the threshold for rejecting the null hypothesis. Both α and β should be as small as possible since they represent error probabilities, though they are rarely zero.

The power of a hypothesis test, calculated as 1 minus β, can be increased by increasing the sample size. Higher power means a greater probability of correctly rejecting a false null hypothesis. Researchers aim for power as close to one as possible to minimize the risk of Type II error.

The four outcomes depend on whether the null hypothesis is true or false and your decision to reject or fail to reject it. You can correctly fail to reject a true null hypothesis, incorrectly reject a true null hypothesis (Type I error), incorrectly fail to reject a false null hypothesis (Type II error), or correctly reject a false null hypothesis.

Both α and β represent probabilities of making incorrect decisions in hypothesis testing. Type I error leads to false rejection of a true null hypothesis, while Type II error results in false acceptance of a false null hypothesis. Minimizing both errors ensures more reliable and valid statistical conclusions.