When testing a hypothesis, you must determine if it is a one-tailed or a two-tailed test. The most common format is a two-tailed test, meaning the critical region is located in both tails of the distribution. This is also referred to as a non-directional hypothesis.
This type of test is associated with a "neutral" alternative hypothesis. Here are some examples:
The alternative option is a one-tailed test. As the name implies, the critical region lies in only one tail of the distribution. This is also called a directional hypothesis. The image below shows a right-tailed test. A left-tailed test would be another type of one-tailed test.
This type of test is associated with a more specific alternative claim. Here are some examples: