In most scientific experiments, it is difficult or impossible to prove that something is true. Instead, many scientists put forward hypotheses about what they think is going to happen. Hypotheses can be two or more possibilities that are contradictory, only one can be true, and exhaustive, they cover all possible outcomes. The hypothesis that is held to be true is called the* null hypothesis* and the other hypotheses are called the* alternative hypotheses*.

With a hypothesis, a scientist is trying to explain an event or observation based on current information. Using the hypothesis, predictions can be made and then tested. A good hypothesis is one that explains all aspects of the observation, is the simplest possible explanation, can be expressed so that predictions can be made about it, and finally, is testable through experimentation.

The fact that this hypothesis is held to be true, even temporarily, is what is being tested in the experiments. Often, it states that there is not going to be a change or effect due to the scientific experimentation. During the experiment, the scientist is trying to either reject or fail to reject this hypothesis. By rejecting it, then it follows that one of the alternative hypotheses is correct, or more correct than the null hypothesis.

It is close to impossible to prove or accept something in science. Instead, hypotheses are rejected or failed to be rejected. For example, a null hypothesis may be that a particular drug will have no effect on those people that it is given to. If an effect is seen within the drugged group, then the null hypothesis is rejected in favor of an alternative hypothesis. If no effect seems to occur in the drugged group, then it is failed to be rejected and usually, further tests are needed.

In statistics, the null hypothesis is seen as the hypothesis that has no significant statistical difference. In other words, it is a statement of statistical equality. It doesn’t have to equal a value exactly, but the hypothesis and the observed sample have to be similar, not different, enough to reject the hypothesis. If the null hypothesis is rejected, that means it is significantly different statistically from the observed group and this difference is not due to chance.

When a null hypothesis is not rejected, it is seen as being statistically similar. This similarity is often attributed to a chance sampling error, meaning the amount of difference is due to chance. If it is rejected, this is not a failure on the experimenter’s behalf. Actually, most researchers and scientists have little or no expectation of the hypothesis being true. A rejected null hypothesis is a significant outcome in scientific experiments.