About Scientific method and statistical terms

In science, the term used for an assumption is hypothesis

An assumption that no differences exists is a null-hypothesis

Beware that a hypothesis is not necessarily a scientific hypothesis. The latter must be falsifiable, implying that it must be possible to identify observable data that potentially conflict with predictions deduced from the hypothesis

The statistical principle for stating a null-hypothesis in science is that if observed data conflicts with, or are considerably unlikely, the consequence is to reject the null hypothesis

The p-value (p = probability) is a measure of how unlikely the observed data are, if we adhere to a null-hypothesis. Common probabilities used in science to reject a null hypothesis are e.g., 0.05, (that is 5%), 0.01, (that is 1%) and 0.001 (which is 0.1%). The numbers are therefore often termed as significance levels