A hypothesis test is either:
- One-tailed or two-tailed
A one-tailed test can be either lower-tailed or upper-tailed.
- One-sample or two-sample
- Paired or unpaired
| Term | Description |
|---|---|
| One-sample test | Uses one test sample. |
| One-tailed test | Rejection region is the lower tail or the upper tail of the sampling distribution under the null hypothesis H0. |
| Lower-tailed test | Alternate hypothesis (H1): μ < μ0 |
| Upper-tailed test | Alternate hypothesis (H1): μ > μ0 |
| Two-tailed test | The null hypothesis assumes that μ = μ0 where μ0 is a specified value. Two-tailed test considers both lower and upper tails of distribution of test statistic. Alternate hypothesis (H1): μ ≠ μ0 |
| Two-sample test | Uses two test samples. |
| Paired test | Compares study subjects at two different times. The null and alternative hypotheses are the same as one sample test. The paired test becomes a one-sample test because the test considers the differences between sample values before and after the subjects are exposed to treatment. |
| Unpaired test | Compares different subjects drawn from two independent populations. H0): μ1 = μ2 The alternate hypotheses are as follows:
|