When carrying out statistical hypothesis testing, there are two rival hypotheses (or assertions) known as the null hypothesis and the alternative hypothesis. They may be used when testing different ideas and assertions when carrying out scientific research or testing in order to find an accurate outcome based on facts.

The null hypothesis will generally correspond to a default or general assertion when carrying out testing. When carrying out tests, every single assertion needs to be falsified against a number of observed data pieces. This is done by collecting together a series of possible outcomes or a set of data and then measuring how probable each of which is true. For example, you could say that when looking at two phenomena that can be measured, the null hypothesis is that there is no relationship between the two. If you are looking at how a drug responds to a specific disease, the null hypothesis is that the drug has had no effect at all.

The alternative hypothesis may be described as the negation or opposite of the null hypothesis. In the above examples, an alternative hypothesis would be that there is a relationship between the two phenomena that can be measured. Likewise, the drug will have an effect on treating the disease.

These two hypotheses are then compared using statistical testing. During this testing, the scientist will not try to prove their alternative hypothesis is true, they will try to prove the null hypothesis wrong. This is where the name ‘null’ comes from as the testing may help ‘nullify’ the null hypothesis.

The null hypothesis is a term that was invented by an English geneticist and statistician called Ronald Fisher. Although this term is generally paired with the alternative hypothesis, this was not actually developed by Fisher and was actually introduced by Jerzy Neyman and Egon Pearson.

The null hypothesis will generally correspond to a default or general assertion when carrying out testing. When carrying out tests, every single assertion needs to be falsified against a number of observed data pieces. This is done by collecting together a series of possible outcomes or a set of data and then measuring how probable each of which is true. For example, you could say that when looking at two phenomena that can be measured, the null hypothesis is that there is no relationship between the two. If you are looking at how a drug responds to a specific disease, the null hypothesis is that the drug has had no effect at all.

The alternative hypothesis may be described as the negation or opposite of the null hypothesis. In the above examples, an alternative hypothesis would be that there is a relationship between the two phenomena that can be measured. Likewise, the drug will have an effect on treating the disease.

These two hypotheses are then compared using statistical testing. During this testing, the scientist will not try to prove their alternative hypothesis is true, they will try to prove the null hypothesis wrong. This is where the name ‘null’ comes from as the testing may help ‘nullify’ the null hypothesis.

The null hypothesis is a term that was invented by an English geneticist and statistician called Ronald Fisher. Although this term is generally paired with the alternative hypothesis, this was not actually developed by Fisher and was actually introduced by Jerzy Neyman and Egon Pearson.