One-sided significance test
Write H₀, find the rejection region at the 5 % level, judge the type 1 error and read the result correctly.
- France: grade 11
- Germany: grade 12
- Syria: grade 12
- Palestine: grade 11
Guide
In a test a null hypothesis H₀ about the probability p is set up, for example H₀: p ≤ 0.2 (right-sided test) against H₁: p > 0.2. You take a sample of size n and count the successes X.
The rejection region holds the values of X that are very unlikely if H₀ is true: its probability is at most the significance level α (for example 5 %). You find the border with the cumulative table of the binomial distribution.
Type 1 error: H₀ is rejected although H₀ is true (probability at most α). Type 2 error: H₀ is not rejected although H₁ is true. A test never proves H₀: «H₀ not rejected» only means that the sample does not speak against H₀.