What are the differences between a regular hypothesis, null hypothesis, and alternative hypothesis?
The Null and Alternative Hypotheses are statistical terms related to a specific binary test. A “regular” hypothesis may be any supposition or conjecture devised to explain observed phenomena.
The critical difference is that the common term “hypothesis” can be a very general opinion in common usage, while the specific null and alternative hypothesis have strict definitions applied to the analysis of statistical data. The “null hypothesis” is a condition that you wish to disprove on the basis of statistical analysis. The “alternative hypothesis” must be the opposite of the null hypothesis. If the null and alternative hypotheses are not opposites, you do NOT have a valid statistical test hypothesis pair.
For example, a regular hypothesis may be something like, “Most people like colored chairs.”. A null hypothesis might be, “More than 50% of colored chairs are blue.” The corresponding alternative hypothesis MUST be: “More than 50% of colored chairs are not blue.” It cannot be a different statement such as, “More than 50% of colored chairs are green.” It must be the opposite of the null hypothesis. If you want to test a different condition, you need to construct a different null/alternative hypothesis pair.